trip based model

Table removed due to copyright \ restrictions.

Table removed due to copyright restrictions.

Open Access is an initiative that aims to make scientific research freely available to all. To date our community has made over 100 million downloads. It’s based on principles of collaboration, unobstructed discovery, and, most importantly, scientific progression. As PhD students, we found it difficult to access the research we needed, so we decided to create a new Open Access publisher that levels the playing field for scientists across the world. How? By making research easy to access, and puts the academic needs of the researchers before the business interests of publishers.

We are a community of more than 103,000 authors and editors from 3,291 institutions spanning 160 countries, including Nobel Prize winners and some of the world’s most-cited researchers. Publishing on IntechOpen allows authors to earn citations and find new collaborators, meaning more people see your work not only from your own field of study, but from other related fields too.

Brief introduction to this section that descibes Open Access especially from an IntechOpen perspective

Want to get in touch? Contact our London head office or media team here

Our team is growing all the time, so we’re always on the lookout for smart people who want to help us reshape the world of scientific publishing.

Home > Books > Models and Technologies for Smart, Sustainable and Safe Transportation Systems

Recent Progress in Activity-Based Travel Demand Modeling: Rising Data and Applicability

Submitted: 18 May 2020 Reviewed: 31 August 2020 Published: 17 September 2020

DOI: 10.5772/intechopen.93827

Cite this chapter

There are two ways to cite this chapter:

From the Edited Volume

Models and Technologies for Smart, Sustainable and Safe Transportation Systems

Edited by Stefano de Luca, Roberta Di Pace and Chiara Fiori

To purchase hard copies of this book, please contact the representative in India: CBS Publishers & Distributors Pvt. Ltd. www.cbspd.com | [email protected]

Chapter metrics overview

1,061 Chapter Downloads

Impact of this chapter

Total Chapter Downloads on intechopen.com

Total Chapter Views on intechopen.com

Over 30 years have passed since activity-based travel demand models (ABMs) emerged to overcome the limitations of the preceding models which have dominated the field for over 50 years. Activity-based models are valuable tools for transportation planning and analysis, detailing the tour and mode-restricted nature of the household and individual travel choices. Nevertheless, no single approach has emerged as a dominant method, and research continues to improve ABM features to make them more accurate, robust, and practical. This paper describes the state of art and practice, including the ongoing ABM research covering both demand and supply considerations. Despite the substantial developments, ABM’s abilities in reflecting behavioral realism are still limited. Possible solutions to address this issue include increasing the inaccuracy of the primary data, improved integrity of ABMs across days of the week, and tackling the uncertainty via integrating demand and supply. Opportunities exist to test, the feasibility of spatial transferability of ABMs to new geographical contexts along with expanding the applicability of ABMs in transportation policy-making.

• activity-based models
• travel demand forecasting
• transportation planning
• transferability of transport demand models

Author Information

Atousa tajaddini *.

• Institute of Transport Studies, Monash University, Australia

Geoffrey Rose

Kara m. kockelman.

• The University of Texas, The United States of America

Hai L. Vu *

*Address all correspondence to: [email protected] and [email protected]

1. Introduction

In recent years, behaviorally oriented activity-based travel demand models (ABMs) have received much attention, and the significance of these models in the analysis of travel demand is well documented in the literature [ 1 , 2 ]. These models are found to be consistent and realistic in several fundamental aspects. They possess some significant advantages over the simple aggregated trip-based travel demand models [ 3 ]. To achieve this, ABMs consider the linkage among activities and travel for an individual as well as different people within the same household and place more attention to the constraints of time and space. In other words, these models are capable of integrating both the activity, time, and spatial dimensions. The comprehensive advantages of activity-based models in comparison to the trip-based models have been discussed in previous papers [ 4 , 5 , 6 , 7 , 8 ]. Activity-based models are suitable for a wider variety of transportation policies involving individual decisions such as congestion pricing and ridesharing. More especially, enabling the relationship between activity and behavioral pattern of trip making is one of the main reasons for the shift from the aggregate-level in trip based models to disaggregate-level provided by ABMs [ 9 ].

Activity-based travel demand models (ABMs) can be classified into two main groups: Utility maximization-based econometric models and rule-based computational process models (CPM). Utility maximization-based econometric models apply different econometric structures such as logit, probit, hazard-based, and ordered response models. While the logit models rely on different assumptions about the distribution of the error terms in the utility functions, hazard-based models use the duration of activity based on end-of-duration occurrence to generate activity schedules [ 10 ]. Rule-based computational process models apply different sets of condition-action rules and focus on the implementation of daily travel and ordering activities to mimic individuals’ behavior when constructing schedules. In addition to the aforementioned models, other approaches can be employed either in combination with these models or separately to develop activity-based models. Examples include agent-based and time-space prism approaches. While an agent-based approach allows agents to learn, modify, and improve their interactions with other agents as well as their dynamic environment, time-space prisms are utilized to capture spatial and temporal constraints under which individuals construct the patterns of their activities and trips. Figure 1 exhibits critical elements of ABM such as activity generation, activity scheduling, and mobility choices. It also provides a comparison among the notable existing travel demand models regarding their different elements. The development of activity-based travel demand models has been reviewed comprehensively in previous studies [ 10 , 11 ]. Table 1 provides a summary of the literature on the evolution of these models over time by introducing the notable existing developed models and highlighting their limitations.

Components of activity-based travel demand models.

ABM evolution over time.

Despite the existence of many models as listed in Table 1 , ABM’s abilities in reflecting behavioral realism are still limited [ 40 ]. The capability of ABM models in predicting individual travel movements can be evaluated from two perspectives of input (data) and output (applicability). Activity schedules are an essential input into the ABM model. From an input point of view, the necessity of deriving activity schedules from dynamic resources together with their challenges will be reviewed. From the applicability perspective, the application of ABM output in integration with dynamic traffic assignment (DTA) models, transferring to a new geographical context, and why and how it is applied in transport planning management will also be discussed. To this end, the first part of this paper will review the new real-time data resources revealing the pattern and traces of traveler’s mobility at a large scale and over an extended period of time. The big data enables new ABM models to reflect mobility behavior on an unprecedented level of detail while collecting data over a longer period (e.g., more than one typical day) would improve the behavioral realism in trip making [ 41 ]. The second part of this paper looks into the applicability of ABM models. This part includes (i) gap investigation in enriching ABMs by integrating time-dependent OD matrices produced by ABMs with dynamic traffic assignment; (ii) investigation of ABMs’ applicability in transferring from one region to another; and (iii) enriching the capability of ABMs by moving beyond the transportation domain to other such as environment and management strategies.

The remainder of the paper is organized as follows. Section 2 introduces new data sources such as mobile phone call data records, transit smart cards, and GPS data where the influence of new data sources on the planning of activities, formation, and analysis of the travel behavior of individuals will be investigated. This section also introduces activity-based travel demand models, which generates activity-travel schedules longer than a typical day. Section 3 describes the existing experiences in transferring utility-based and CPM activity-based travel demand models from one geographical area to another. This section also reviews the integration of ABM models with dynamic traffic assignment and other models such as air quality models. The possibility of using activity-based models in travel demand management strategies with a focus on car-sharing and telecommuting are considered as examples. The last section concludes the paper and identifies remaining challenges in the area of activity-based travel demand modeling.

2. ABMs and the emerging of big data

This section provides an overview of the role of big data in replacing the traditional data sources, and the changes in activity-based travel demand models given these newly available data.

2.1 Improvements in activity-based travel demand modeling

It is more than half a century that transportation planners try to understand how individuals schedule their activities and travel to improve urban mobility and accessibility. The evolution of travel demand modeling from trip-based to activity-based highlighted the need for high-resolution databases including sociodemographic and economic attributes of individuals and travel characteristics. Today, with the rapid advancements in computation, technology, and applications, the intelligent transportation systems (ITS) have revolutionized the analysis of travel behavior by having more accurate data, removing human errors, and making use of the vast amount of available data [ 42 ]. Tools such as GPS devices, smartphones, smart card data, and social networking sites all have the potential to track the movements and activities of individuals by recording and retaining the relevant data continuously over time. Most of the traditional travel survey data are rich in detail. However, it can result in biased travel demand models because of incomplete self-reports and inaccurate scheduling patterns. Therefore, in this section, the common tools used in collecting big data are introduced and the progress made in the area of extracting big data sources is discussed.

2.1.1 Cell phone data

A call detail record (CDR) is a data record produced by a telephone exchange and consists of spatiotemporal information on the recent system usage [ 40 ], which can track people’s movements. This CDR data can be processed and applied in activity-based travel demand modelings to better understand human mobility and obtain more accurate origin-destination (OD) tables [ 43 ]. The first attempt using CDR data was a study of Caceres et al. [ 44 ], who applied mobile phone data to generate OD matrices. Their concept was then formalized by Wang et al. [ 45 ] to obtain transient OD matrices by counting trips for each pair of the following calls from two different telephone (cell) towers at the same hour. Afterward, using the shortest path algorithm, OD trips are assigned to the road network. In the area of urban activity recognition, Farrahi et al. [ 46 ] applied two probabilistic methods (i.e., Latent Dirichlet Allocation (LDA) and Author Topic Models, ATM) to cluster CDR trajectories according to their temporal aspects to discover the home and work activities. Considering the spatial aspect of CDR data, Phithakkitnukoon et al. [ 47 ] applied auxiliary land use data and geographical information database to find possible activities around a certain cell tower. And considering both the temporal-spatial aspect of CDR, Widhalm et al. [ 48 ] used an undirected relational Markov network to infer urban activities. They extracted activity patterns for Boston and Vienna by analyzing cell phone data (activity time, duration, and land use). Their results show that trip sequence patterns and activity scheduling observed from datasets were compatible with city surveys as well as the stability of generated activity clusters across time. In a more recent study, [ 49 ] an unsupervised generative state-space model is applied to extract user activity patterns from CDR data. Furthermore, it has been shown that the method of CDR sampling is as significant as survey sampling. For example, in one study [ 50 ], CDR and survey data is used during a period of six months to investigate the daily mobility for Paris and Chicago. The result shows that 90% of travel patterns observed in both surveys are compatible with phone data. In another similar study [ 51 ], a probabilistic induction was proposed using motifs (daily mobility network), time of day activity sequence, and land use classification to produce activity types. CDR data of Singapore was used by Jiang et al. [ 52 ] to produce activity-based human mobility patterns.

In the context of activity-based transport modeling, Zilske et al. [ 53 ] replaced travel diaries with CDRs as input data for agent-based traffic simulation. They first generated the synthetic CDR data, then the MATSim simulation software was used to identify every observed person as an agent to convert call information into activity. They fused the CDR data set with traffic counts in their next paper [ 54 ], to reduce the Spatio-temporal uncertainty.

In summary, the findings reported from different studies indicated the major implications of mobile phone records on the estimation of travel demand variables including travel time, mode and route choice as well as OD demand and traffic flow estimation; however; in practice, the information generated from CDR data are yet to be used widely in simulation models. This is mainly because of the conflict between either level of resolution or format and completeness of model and data [ 55 ].

2.1.2 Smart card data

Smart card systems with on- and off-boarding information gained much popularity in large public transport systems all over the world, and have become a new source of data to understand and identify the Spatio-temporal travel patterns of the individual passengers. The smart card data are investigated in various studies such as activity identification, scheduling, agent-based transport models, and simulation [ 56 ]. Besides, in other studies [ 57 , 58 , 59 ] smart card data was used as an analysis tool in investigating the passenger movements, city structure, and city area functions. Similarly, in the recent study [ 60 ], a visual analysis system called PeopleVis was introduced to examine the smart card data (SCD) and predict the travel behavior of each passenger. They used one-week SCD in the city of Beijing and found a group of “familiar strangers” who did not know each other but had lots of similarities in their trip choices. Zhao et al. [ 61 ] also investigated the group behavior of metro passengers in Zhechen by applying the data mining procedure. After extracting patterns from smart card transaction data, statistical-based and clustering-based methods were applied to detect the passengers’ travel patterns. The results show that a temporally regular passenger is very probable to be a spatially regular passenger. The disaggregated nature of smart card data represents suitable input to multi-agent simulation frameworks. For example, the smart card data is used to generate activity plans and implement an agent-based microsimulation of public transport in two cities of Amsterdam and Rotterdam [ 62 ]. An agent-based transport simulation is developed for Singapore’s public transport using MATSim environment [ 63 ]. Unlike Bouman’s study, they considered the interaction of public transport with private vehicles. The study of Fourie et al. [ 64 ] was another research work to present the possibility of integrating big data algorithms with agent-based transport models. Zhu [ 65 ] compared one-week transaction data of smart cards in Shanghai and Singapore. They found feasibility in generating continuous transit use profiles for different types of cardholders. However, to have a better understanding of the patterns and activity behaviors, in addition to collecting the data from smart cards, one should integrate them with other data set.

2.1.3 GPS data

In travel demand modeling, it is important to have accurate and complete travel survey data including trip purpose, length, and companions, travel demand, origin and destination, and time of the day. Since the 1990s, the global positioning system (GPS) became popular for civil engineering applications, especially in the field of transportation as it provides a means of tracking some of the above variables. In the literature, methods of processing the GPS data and identifying activities can be classified according to different approaches such as rule-based and Bayesian model [ 66 ]; fuzzy logic [ 67 ]; multilayer perceptron [ 68 ]; and support vector machine learning [ 69 ]. Nevertheless, the disadvantages of using GPS data include the cost, sample size limitation, and the need to retrieve and distribute GPS devices to participate. Since smartphones are becoming one of the human accessories while equipped with a GPS module, they can be considered as a replacement of the GPS device to gather travel data. In this regard, CDR from smartphones is used [ 70 ] to estimate origin-destination matrices, or a smartphone-based application is used [ 71 ] to map the semiformal minibus services in Kampala (Uganda) and to count passenger boarding and alighting [ 72 ]. In the Netherlands, the Mobidot application is developed for analyzing the mobility patterns of individuals. To deduce travel directions and modes, this application uses the real-time data gathered by sensors of smartphones including GPS, accelerometer, and gyroscope sensors to compare them with existing databases [ 73 ].

Applying smartphones as a replacement of GPS however, holds several restrictions including the draining of smartphone battery and it is not possible to record travel mode and purpose.

2.1.4 Social media data

Today transport modelers, planners, and managers have started to benefit from the popularity of social networking data. There are different kinds of social media data such as Twitter, Instagram, and LinkedIn data, which consist of normal text, hash-tag (#), and check-in data. As hash-tag and check-in data are related to an activity, location or event, they can be used as meaningful resources in analysis of destination/origin of the activity [ 74 ]. According to the literature, social media has a great influence on different aspects of travel demand modeling [ 75 ]. Using social media instead of traditional data collection methods was investigated in different studies [ 76 ]. The way of processing these data to extract useful information is challenging as investigated in different studies [ 77 , 78 ]. Various studies [ 79 , 80 , 81 , 82 ] also examined social media data to understand the mobility behavior of a large group of people. Testing the possibility of evaluating the origin-destination matrix based on location-based social data was researched [ 83 ] or in another similar studies [ 84 , 85 ] where Twitter data was used to estimate OD matrices. The comparison between this new OD with the traditional values produced by the 4-step model proved the great potential of using social media data in modeling aggregate travel behavior. Social media data can be used in other areas such as destination choice modeling [ 86 ], recognizing activity [ 87 ], understanding the patterns of choosing activity [ 80 , 88 , 89 ], and interpreting life-style behaviors via studying activity-location choice patterns [ 90 ].

2.2 Dynamic ABM using a multi-day travel data set

Most existing travel demand modelers have applied the household survey data during the period of one day to construct activity schedules. However, longer periods such as one week or one month gained substantial importance during recent years. For simulating everyday travel behavior and generating schedules, a one-week period provides more comprehensive coverage because it includes weekdays and weekends and represents the weekly routines of individuals in making trips. Periods longer than one week can further provide detail on personal behavior as well as various usage of modes in different ways. So far only a few travel demand models covered a typical week as a studied period. For example rhythm in activity-travel behavior based on the capacity of one week was presented by applying a Kuhn-Tucker method [ 41 ]. Few works have been concentrating on the generation of multiple-day travel dataset. For example, by using large data and surveys, Medina developed two discrete choice models for generating multi-day travel activity types based on the likeliness of the activity [ 91 ]. a sampling method based on activity-travel pattern type clustering [ 92 ] was proposed to extract multi-day activity-travel data according to single-day household travel data. The results show similarities in distributions of intrapersonal variability in multi-day and single-day. MATSim is a popular agent-based simulation for ABM research [ 93 , 94 ], however, it is not appropriate for modeling the multi-day scenarios because MATSim uses the co-evolutionary algorithm to reach the user equilibrium which is a time consuming particularly for multi-day plans. To solve these problems, Ordonez [ 95 ] proposed a differentiation between fixed and flexible activities. Based on different time scales, Lee examined three levels of travel behavior dynamics, namely micro-dynamics (24 hours), macro-dynamics (lifelong travel behavior), meso-dynamics (weekly/monthly/yearly basis) by applying different statistical models [ 96 ]. A learning day-by-day module in another agent-based simulation software SimMobility is proposed [ 97 ]. Furthermore, ADAPTS is one of the few activity-based travel demand models which depends on activity planning horizon data for a longer period than one day, for example, one week or one month [ 33 ].

As highlighted by the above literature review, applying one-day observation data in travel demand modeling provides an inadequate basis of understanding of complex travel behavior to predict the impact of travel demand management strategies. So multi-day data are needed to refine this process. Previously, it was not easy to collect multi-day data, however, today thanks to advantages to technology it is possible to extract data from GPS, smartphones, smart cards, etc. with no burden for the respondent. Models built based on GPS data have been found to be more accurate and precise due to having fewer measurement errors. Collecting call detail records from mobile phones provide modelers with large trip samples and origin-destination matrices, while smart card data are more useful in terms of validation.

3. ABM transferability

We now turn to the recent advances and ongoing research in ABM focused on testing and enhancing geographical transferability and capacity to predict a broader range of impacts than flows and performance of the transport network.

3.1 ABM transferability from one geographical context to another

The spatial transferability of a travel demand model happens when the information or theory of a developed model of one region is applied to a new context [ 98 ]. Transferability can be used not only as a beneficial validation test for the models but also to save the cost and time required to develop a new model. Validation of a model by testing spatial transferability beside other various methods such as base-year and future-year data set is a test of validity which represents the capability of activity-based models in predicting travel behavior in a different context [ 99 ]. The exact theoretical basis and behavioral realism of activity-based travel demand model make them more appropriate for geographic transferability in comparison to traditional trip-based models [ 100 ]. Testing the transferability of ABM was first investigated by Arentze et al. [ 101 ]. They examined the possibility of transferring the ALBATROSS model at both individual and aggregate levels for two municipalities (Voorhout and Apeldoorn) in the Netherlands by simulating activity patterns. The results were satisfactory except for the transportation mode choice. In the United States, the CT-RAMP activity-based model which was developed for the MORPC region then transferred to Lake Tahoe [ 102 ]. In another study, one component of the ADAPTS model showed the potential for having good transferability properties [ 31 ]. The transferability of the DaySim model system developed for Sacramento to four regions in California and two other regions in Florida was investigated in [ 103 ]. The results show that the activity generation and scheduling models can be transferred better than mode and location choice models. The CEMDAP model developed for Dallas Fort Worth (DFW) region was transferred to the southern California region [ 104 ]. Outside the U.S., the TASHA model system developed for Toronto was transferred to London [ 105 ], and also in another study [ 106 ] the transferability of TASHA to the context of the Island of Montreal was assessed. Activity generation, activity location choice, and activity scheduling were three components of TASHA that transferred from Toronto to Montreal. In general, TASHA provided acceptable results at (macro and meso-level) for work and school activities even in some cases better results for Montreal in comparison to Toronto area. The possibility of developing a local area activity-based transport demand model for Berlin by transferring an activity generation model from another geographical area (Los Angeles) and applying the traffic counts of Berlin was investigated [ 107 ]. In their research, the CEMDAP model was applied to achieve a set of possible activity-travel plans, and the MATSim simulation was then used to generate a representative travel demand for the new region. The results were quite encouraging, however, the study indicated a need for further evaluation. In one recent study [ 108 ], an empirical method was used to check the transferability of ABMs between regions. According to their investigations, the most difficult problems with transferability caused by parameters of travel time, travel cost, land use, and logsum accessibilities. They suggested that in the transferability of the ABM from another region, agencies should be aware of finding a region within the same state or with similar urban density, or preferably both in order to improve the results. The possibility of transferring the FEATHERS model to Ho Chi Minh in Vietnam is investigated [ 109 ]. FEATHERS initially is developed for Flanders in Belgium. After calibration of FEATHERs sub-models, testing results using different indicators confirmed the success of transferring the FEATHER’s structure to the new context.

At the theoretical level, a perfect transferable model contributes to the transferability of its underlying behavioral theory, model structure, variable specification and coefficient to the new context. However, perfect transferability is not easy to achieve due to different policy and planning needs as well as the size of the regions, and the availability of data and other resources. Although the results of several transferred ABM model systems seem to have worked reasonably, it is equally important to assess how much accuracy is important in transferring models and how best and where to transfer models from.

3.2 ABM transferability to other non-transport domain

One of the advantages of the activity-based travel demand models over trip-based models is its capability to generate various performance indicators such as emission, health-related indicators, social exclusion, well-being, and quality of life indicators. Application of disaggregate models for the area of emission and air quality analysis was introduced by Shiftan [ 110 ] who investigated the Portland activity-based model in comparison to trip-based models. In another study [ 111 ], the same author integrated the Portland activity-based model with MOBILE5 emission model to study the effects of travel demand techniques on air quality. Regarding the integration of ABM with the emission model, the Albatross ABM model was coupled with MIMOSA (macroscopic emission model) [ 112 ] considering the usage of fuel and the amount of produced emission as a function of travel speed. A study in [ 113 ] added one dispersion model (AUROTA) to the previous integration of Albatross and MIMOSA to predict the hourly ambient pollutant. Albatross linked with a probabilistic air quality system was employed [ 114 ] in air quality assessment study. TASHA was another activity-based model, which has been extensively employed in air quality studies. For example, this model was integrated [ 29 , 115 ] with MOBILE6.2 to quantify vehicle emissions in Toronto. In their study, EMME/2 was used in the traffic assignment part. The previous research was improved [ 116 ] by replacing EMME/2 with MATSim as an agent-based DTA model. This TASHA-MATSim chain was used in the research [ 117 ] with the integration of MOBILE6.2C (emission model) and CALPUFF (dispersion model). OpenAMOS linked with MOVES emission model [ 118 ], and ADAPTS linked with MOVES [ 119 ] together with Sacramento ABM model [ 120 ] are among recent studies which represented the application of activity-based models in analyzing the impacts of vehicular emissions.

Human well-being and personal satisfaction play an important role in social progression [ 121 ]. To understand the theory behind human happiness, transport policies concentrated on the concept of utility as a tool to increase activity, goods, and services [ 122 , 123 ]. The issue of well-being as a policy objective is addressed in the literature and measured through various indicators, which show personal satisfaction and growth. For example, in the study by Hensher and Metz [ 124 , 125 ], saving time which leads to engagement in more activities was introduced as one of the benefits of measuring transport performance. Spatial accessibility was another benefit of travel that provides a range of activities that can be reasonably reached by individuals [ 126 ]. A dynamic ordinal logit model was developed [ 127 ] based on the collected data on happiness for a single activity in Melbourne. The authors found different activity types, which have different influences on the happiness that each individual experienced. Well-being can be integrated into activity-based models based on random utility theory. In terms of modeling, a framework was introduced [ 122 ] considering well-being data to improve activity-based travel demand models. According to their hypothesis, well-being is the final aim of activity patterns. They applied a random utility framework and considered well-being measures as indicators of the utility of activity patterns, and planned to test their framework empirically by adding well-being measurement equations to the DRCOG’s activity-based model.

The above literature review showed the importance of applying traffic models to generate input data to other models such as the air quality model. The accuracy of emission models is highly dependent on the level of detail in transport demand model inputs. Activity-based and agent-based models are supposed to describe reality more accurately by providing more detailed traffic data. Beyond measurement of air quality, well-being and health have drawn increasing attention. The health impact of changes in travel behavior, health inequalities, and social justice can be assessed within the activity-based platform [ 128 ]. With the help of geospatial data acquisition technologies like GPS, behavioral information with health data can be integrated into the development of an activity-based model to provide policies that affect the balance of transport and well-being.

3.3 ABM integration with dynamic traffic assignment

In parallel with the travel demand modeling, on the supply side, the conventional supply models used to be static, which import constant origin-destination flows as an input and produce static congestion patterns as an output. Consequently, these models were unable to represent the flow dynamics in a clear and detailed manner. Dynamic traffic assignment (DTA) models have emerged to address this issue and are capable of capturing the variability of traffic conditions throughout the day. It is evident that the shift of analysis from trips to activities in the demand modeling, as well as, the substitution of the static traffic assignment with dynamic traffic assignment in the supply side, can provide more realistic results in the planning process. Furthermore, the combination of ABM and DTA can better represent the interactions between human activity, their scheduling decision, and the underlying congested networks. Nevertheless, according to the study of [ 11 ], the integration of ABM with DTA received little attention and still requires further theoretical development. There are different approaches to the integration of ABM and DTA, which started with a sequential integration. In this type of integration, exchanging data between two major model components (ABM and DTA) happens at the end of the full iteration, to generate daily activity patterns for all synthetic population in an area of study, the activity-based model is run for the whole period of a complete day. The outputs of the ABM model which are lists of activities and plans are then fed into the DTA model. The DTA model generates a new set of time-dependent skim matrices as inputs to ABM for the next iteration. This process is continued until the convergence will be reached in the OD matrices output. Model systems applying the sequential integration paradigm can be found in most of the studies in the literature. For example, Castiglione [ 129 ] integrated DaySim which is an activity-based travel demand model developed for Sacramento with a disaggregate dynamic network traffic assignment tool TRANSIMS router. Bekhor [ 130 ] investigated the possibility of coupling the Tel Aviv activity-based model with MATSim as an agent-based dynamic assignment framework. Hao [ 116 ] integrated the TASHA model with MATSim. Ziemke [ 107 ] integrated CEMDAP, which is an activity-based model with MATSim to check the possibility of transferring an activity-based model from one geographic region to another. Lin [ 131 ] introduced the fixed-point formulation of integrated CEMDAP as an activity-based model with an Interactive System for Transport Algorithms (VISTA). Based on the mathematical algorithm of household activity pattern problem (HAPP), ABM and DTA were integrated [ 132 ] by presenting the dynamic activity-travel assignment model (DATA) which is an integrated formulation in the multi-state super network framework.

In the sequential integration, the ABM and DTA models run separately until they reach convergence. At the end of an iteration, these models perform data exchange before iterate again. Therefore, this kind of integrated framework cannot react quickly and positively to network dynamics and is unable to adapt to real-time information available to each traveler. In addressing this limitation, integrated models that adopt a much tighter integration framework have been developed recently. This approach is quite similar to the sequential approach, however; the resolution of time for ABM simulation is one minute rather than 24 hours (complete day). Relating to this level of dynamic integration, Pendyala [ 133 ] investigated the possibility of integrating OpenAMOS which is an activity-travel demand model with DTA tool name MALTA (Multiresolution Assignment and loading of traffic activities) with appropriate feedback to the land-use model system. For increasing the level of dynamic integration of ABM and DTA models, dynamic integration having pre-trip enroute information with full activity-travel choice adjustments has been introduced. In this level of ABM & DTA integration, it is assumed that pre-trip information is available for travelers about the condition of the network. It means that travelers are capable of adjusting activity-travel choices since they have access to pre-trip and Enroute travel information. Another tightly integrated modeling framework was proposed in [ 134 ] to integrate ABM (openAMOS) and DTA (DTALite) to capture activity-travel demand and traffic dynamics in an on-line environment. This model is capable of providing an estimation of traffic management strategies and real-time traveler information provision. Zockaie et al. [ 135 ] presented a simulation framework to integrate the relevant elements of an activity-based model with a dynamic traffic assignment to predict the operational impacts related to congestion pricing policies. Auld et al. [ 38 ] developed an agent-based modeling framework (POLARIS) which integrates dynamic simulation of travel demand, network supply, and network operations to solve the difficulty of integrating dynamic traffic assignment, and disaggregate demand models. A summary of the current literature on ABM and DTA integration is presented in Table 2 .

A summary of the empirical literature on ABM and DTA integration.

The above discussion illustrates that most of the model integration platforms between ABM + DTA work based on sequential integration. This loose coupling platform is the most straightforward and popular approach albeit is not responsive to network short-term dynamics and real-time information. Efforts to develop a comprehensive simulation model that can account for all components of dynamic mobility and management strategies continue. Further developments will have to deal with the implementation of an integrated ABM + DTA platform on a large network to support decision-makers, focus on the integration between activity-based demand models and multimodal assignment [ 143 ] as well as reducing computational efforts via better data exchange procedure and improving model communication efficiency. Defining practical convergence criteria is another issue which needs further investigations. Fully realistic convergence is normally never happened in sequential integration due to applying a pre-defined number of feedback loops in order to save model runtime.

3.4 ABM and travel demand management applications

Travel demand management (TDM) strategies are implemented to increase the efficiency of the transportation system and reduce traffic-related emissions. Some examples include mode shift strategies (encouraging people to use public transport) [ 144 ], time shift (to ride in off-peak hours, congestion pricing), and travel demand reduction [ 145 ] (using shared mobility service or teleworking). Shared transport services including car sharing, bike sharing, and ridesharing have been implemented in most of the transport planning systems across the world. Applying activity-based travel demand models to study the optimal fleet size can be found in different studies in the literature [ 146 , 147 ]. Parking price policies and their impacts on car sharing were investigated using MATSim in [ 148 ]. Results show shared vehicles use more efficient parking spaces in comparison to private vehicles. In the first attempt to model car sharing on more than one typical day [ 149 ] the agent-based simulation (mobitopp) was extended with a car-sharing option to study the travel behavior of the population in the city of Stuttgart in one week. In the recent study of [ 150 ], car sharing was integrated into an activity-based dynamic user equilibrium model to show the interaction between the demand and supply of car sharing. Among all the TDM strategies, telecommuting can be implemented in a shorter time [ 151 , 152 , 153 ]. The results of these studies present a reduction in vehicle-kilometers-traveled (VKT) during peak hours mainly because telecommuters change their trip timetable during these times. This plan rescheduling is also investigated and addressed in different studies [ 154 ] based on the statistical analysis of worker’s decisions about choice and frequency of telecommuting. While the plan rescheduling leads to reducing commute travel, the overall impacts of telecommuting on the formation of worker’s daily activity-travel behavior is challenging. For example, this policy reduced total distance traveled by 75% on telecommuting days while telecommuting could reduce the total commute distance up to 0.8% and 0.7% respectively [ 151 , 155 ]. Based on the adoption and frequency of telecommuting, a joint discrete choice model of home-based commuting was developed for New York city using the revealed preference (RP) survey [ 156 ]. Their results show a powerful relationship among individuals’ attributes, households’ demographics, and work-related factors, and telecommuting adoption and frequency decisions. A similar study [ 157 ] estimated the telecommuting choice and frequency by using a binary choice model and ordered-response model respectively. In terms of using activity-based modeling, [ 158 ] POLARIS activity-based framework was applied to research telecommuting adoption behavior and apply MOVES emission simulator model to assess the consequences of implementing this policy on air quality. Their results show that considering 50% of workers in Chicago with flexible working time hours in comparison to the base case with 12% flexible time hour workers, telecommuting can reduce Vehicle Mile Traveled (VMT) and Vehicle Hour Traveled (VHT) by 0.69% and 2.09% respectively. This policy reduces greenhouse gas by up to 0.71% as well. Pirdavani et al. [ 159 ] investigated the impact of two TDM scenarios (increasing fuel price and considering teleworking) on traffic safety. In this work, FEATHERS model, which is an activity-based model, was applied to produce exposure matrices to have a more reliable assessment. The results show the positive impacts of two scenarios on safety ( Figure 2 ).

Travel demand management policies within the activity-based platform.

The above section explores the relationship between transport demand management policies and travel behavior in the ABM context. The use of an activity-based travel demand model provides flexibility to employ a range of policy scenarios, and at the same time, the results are as detailed as possible to obtain the impact of policies on a disaggregated level. The finding highlights the importance of implementing different transportation policies management together to reach the most appropriate effect in terms of improving sustainability and the environment. The discussion emphasizes the need for considering more comprehensive transportation and environmental policies concerning sustainability to tackle travel planning in light of the increasingly diverse and complex travel patterns.

4. Summary and research directions

The use of activity-based models to capture complex underlying human’s travel behavior is growing. In this paper, we began by introducing the components of activity-based models and the evolution of the existing developed ABM models. In the first part of this paper, the new resources of data for travel demand analysis were introduced. In the new era of travel demand modeling, we need to deal with a dynamic, large sample, time-series data provided from new devices, and as a result manage observation covering days, weeks, and even months. The outcome of the recent works revealed that since activity-based models originated from the concept of individual travel patterns rather than aggregate flows, they highly suited to these new big data sources. These big datasets, which document human movements, include the information about mobility traces and activities carried out. Based on the in-depth and critical review of the literature, it is clear that while these big datasets provide detailed insight into travel behavior, challenges remain in extracting the right information and appropriately integrating them into the travel demand models. In particular, extracting personal characteristics and trip information like trip purpose and mode of transport are still open problems as these big data resources which provide space-time traces of trip-maker behaviors. Research works along these lines have been started as it was reviewed in the first part; however, further researches should be conducted to handle the uncertainty of big data mobility traces in the modeling process. Also, new methods should be investigated to validate the results for each step of the data analysis and mining. The possibility of fusing data from different available datasets needs further investigation. For instance, to understand the mode inference both data from the smart card and CDRs can be analyzed simultaneously. Another challenging issue regarding the application of this rich new data in transport modeling is that the need for methodologies to extract useful information needed regarding the traveler’s in-home and out-of-home activity patterns, which highlights the combination of data science, soft computing-based approaches, and transport research methods. It requires new Different algorithms such as statistical, genetic, evolutionary, and fuzzy as well as different techniques including advanced text and data mining, natural language processing, and machine learning.

The spatial transferability of activity-based travel demand models remains an important issue. Generally, it is found that the transferability of these models is more feasible than trip-based models, especially between two different regions with similar density or even between two areas in the same state. To date, most of the transferability research in activity-based travel demand modeling is motivated by a desire to save time, and very few studies that applied spatial transferability of activity-based models have undertaken rigorous validation of the results. While literature showed successful model transferability in terms of transferring activity/tour generation, time-of-day choice components, more studies are required on the model transferability regarding mode and location choice models as well as the validation test of activity-based models in different levels, i.e., micro, meso, and macro models.

As part of the second section of this study, this paper reviewed the progress made in the integration of activity-based models with dynamic traffic assignment.

Based on the literature, although evolution has occurred in DTA models, the loose coupling (sequential method) between ABM and DATA models still dominate the field. Two main challenges remain, namely poor convergence quality and excessively long run time. Replacing MATSim as a dynamic traffic assignment tool with other route assignment algorithms in recent years was a technical solution to loose coupling, which considered route choice as another facet of a multi-dimensional choice problem. MATSim provides not only an integration between the demand and supply side, but it can also act as a stand-alone agent-based modeling framework. However; MATSim potential drawbacks include being based on unrealistic assumptions of utility maximization and perfect information. To remove these unrealistic rational behavioral assumptions, applying other approaches such as a new innovative method of behavioral user equilibrium (BUE) is needed. This method helps trip-makers to reach certain utility-level rather than maximize the utility of their trip making [ 160 ]. Work along this approach has started (e.g., [ 161 ]).

The capability of activity-based models in generating other kinds of performance indicators in addition to OD matrices was also reviewed. Literature proved activity-based models generate more detailed results as inputs to air quality models, however; error rises from the accuracy of the information has a relevant impact on the process of integration. So it is necessary to do a comprehensive analysis of the uncertainties in traffic data. Literature proved that despite of the improvements in such disaggregate frameworks and the capability of these models in replicating policy sensitive simulation environment; there is yet to develop the best and perfect traffic-emission-air quality model. While the issue of health has drawn extensive attention from many fields, activity-based travel demand models have proved to have the potential to be used in estimating health-related indicators such as well-being. However, very few studies have been found to investigate the theories required to extend the random utility model based on happiness. While it is proved that mobility and environment have direct impacts on transport-related health [ 162 ], investigations on how travel mode preferences and air pollution exposure are related in this context are needed. Another area of research within ABM platform which is yet to be studied is the relationship between individual exposure to air pollution and mobility, especially in space, and time.

In the last part of this paper, the capability of activity-based models in the analysis of traffic demand management was investigated. Generally, the influence of telecommuting on both travel demand and network operation is still incomplete. Very few studies were found in which activity-based framework is used to simulate the potential impacts of telecommuting on traffic congestion and network operation where the real power of activity-based models lie.

In conclusion, while there are still open problems in activity-based travel demand models, there has been a lot of progress being made which is evidenced by the various recent and on-going researches reviewed in this paper. The review showed that by applying different methodologies in the modeling of different aspects of activity-based models, these models are becoming more developed, robust, and practical and become an inevitable tool for transport practitioners, city planners, and policy decision-makers alike.

Acknowledgments

The research work presented in this paper was supported by the Australian Government-Department of Education under Research Training Program (RTP Stipend) award.

• 1. Vovsha, P., M. Bradley, and J.L. Bowman. Activity-based travel forecasting models in the United States: progress since 1995 and prospects for the future . in at EIRASS Conference on Progress in Activity-Based Analysis . 2005. Vaeshartelt Castle, Maastricht.
• 2. Joe Castiglione, M.B., and John Gliebe, Activity-Based Travel Demand Models A Primer , in SHRP 2 Report S2-C46-RR-1 . 2015, TRANSPORTATION RESEARCH BOARD: Washington, D.C.
• 3. McNally, M.G. and C.R. Rindt, The activity-based approach , in Handbook of Transport Modelling: 2nd Edition . 2007, Emerald Group Publishing Limited. p. 55-73.
• 4. Henson, K., K. Goulias, and R. Golledge, An assessment of activity-based modeling and simulation for applications in operational studies, disaster preparedness, and homeland security. Transportation Letters, 2009. 1 (1): p. 19-39.
• 5. Davidson, B., P. Vovsha, and J. Freedman. New advancements in activity-based models . in Australasian Transport Research Forum . 2011.
• 6. Chu, Z., L. Cheng, and H. Chen, A Review of Activity-Based Travel Demand Modeling , in CICTP 2012: Multimodal Transportation Systems—Convenient, Safe, Cost-Effective, Efficient . 2012. p. 48-59.
• 7. Zhong, M., et al., A comparative analysis of traditional four-step and activity-based travel demand modeling: a case study of Tampa, Florida. Transportation Planning and Technology, 2015. 38 (5): p. 517-533.
• 8. Subbarao, S.S. and K. Krishnarao, Activity based approach to travel demand modelling: An overview. EUROPEAN TRANSPORT-TRASPORTI EUROPEI, 2016 (61).
• 9. Daisy, N.S.M., Hugh; Liu, Lei, Individuals’ Activity-Travel Behavior in Travel Demand Models: A Review of Recent Progress , in 18th COTA International Conference of Transportation Professionals . 2018: Beijing, China.
• 10. Pinjari, A.R. and C.R. Bhat, Activity-based travel demand analysis. A Handbook of Transport Economics, 2011. 10 : p. 213-248.
• 11. Rasouli, S., and Harry Timmermans, Activity-Based Models of Travel Demand: Promises, Progress and Prospects. International Journal of Urban Sciences 2014. 18(1) : p. 31-60
• 12. Lenntorp, B., Paths in space-time environment: A time geographic study of possibilities of individuals. The Royal University of Lund, Department of Geography. Lund Studies in Geography, Series B. Human Geography, 1976. 44 .
• 13. Jones, P., et al., Understanding Travel Behavior Gower Publishing Co. Ltd. Aldershot, UK Google Scholar, 1983.
• 14. Huigen, P.P., Binnen of buiten bereik?: Een sociaal-geografisch onderzoek in Zuidwest-Friesland . 1986: na.
• 15. Dijst, M. and V. Vidakovic, INDIVIDUAL ACTION SPACE IN THE CITY. Activity-based approaches to travel analysis, 1997.
• 16. Kwan, M.-P., GISICAS: AN ACTIVITY-BASED TRAVEL DECISION SUPPORT SYSTEM USING A GIS-INTERFACED COMPUTATIONAL-PROCESS MODEL. Activity-based approaches to travel analysis, 1997.
• 17. Bradley, M.A., Portland Metro, J. L. Bowman, A system of activitybased models for Portland, Oregon , in USDOT report number DOT-T-99-02 . 1998: Washington, D.C.
• 18. Bradley, M., et al. Estimation of an activity-based micro-simulation model for San Francisco . in 80th Annual Meeting of the Transportation Research Board, Washington DC . 2001.
• 19. Vovsha, P., E. Petersen, and R. Donnelly, Microsimulation in travel demand modeling: Lessons learned from the New York best practice model. Transportation Research Record: Journal of the Transportation Research Board, 2002(1805): p. 68-77.
• 20. Consult, P.B., The MORPC travel demand model: Validation and final report. Prepared for the Mid-Ohio Region Planning Commission, 2005.
• 21. Bowman, J. and M. Bradley, Activity-based travel forecasting model for SACOG: Technical memos numbers 1-11. Available at* HYPERLINK" http://jbowman.net "* http://jbowman.net , 2005.
• 22. Bradley, M. and J. Bowman. A summary of design features of activity-based microsimulation models for US MPOs . in White Paper for the Conference on Innovations in Travel Demand Modeling, Austin, TX . 2006.
• 23. Bhat, C., et al., Comprehensive econometric microsimulator for daily activity-travel patterns. Transportation Research Record: Journal of the Transportation Research Board, 2004(1894): p. 57-66.
• 24. Pinjari, A.R., et al., Activity-based travel-demand analysis for metropolitan areas in Texas: CEMDAP Models, Framework, Software Architecture and Application Results . 2006.
• 25. Pendyala, R., et al., Florida activity mobility simulator: overview and preliminary validation results. Transportation Research Record: Journal of the Transportation Research Board, 2005(1921): p. 123-130.
• 26. Davidson, W., et al. CT-RAMP family of activity-based models . in Proceedings of the 33rd Australasian Transport Research Forum (ATRF) . 2010.
• 27. Arentze, T. and H. Timmermans, Albatross: a learning based transportation oriented simulation system . 2000: Citeseer.
• 28. Arentze, T.A. and H.J. Timmermans, A learning-based transportation oriented simulation system. Transportation Research Part B: Methodological, 2004. 38 (7): p. 613-633.
• 29. Miller, E. and M. Roorda, Prototype model of household activity-travel scheduling. Transportation Research Record: Journal of the Transportation Research Board, 2003(1831): p. 114-121.
• 30. Roorda, M.J. and E.J. Miller, Strategies for resolving activity scheduling conflicts: an empirical analysis. 2005.
• 31. Auld, J. and A. Mohammadian, Framework for the development of the agent-based dynamic activity planning and travel scheduling (ADAPTS) model. Transportation Letters, 2009. 1 (3): p. 245-255.
• 32. Auld, J.A. and A. Mohammadian. Planning Constrained Destination Choice Modeling in the Adapts Activity-Based Model . in Proceedings of the Transportation Research Board 90th Annual Meeting, Washington, DC, USA . 2011.
• 33. Auld, J. and A.K. Mohammadian, Activity planning processes in the Agent-based Dynamic Activity Planning and Travel Scheduling (ADAPTS) model. Transportation Research Part A: Policy and Practice, 2012. 46 (8): p. 1386-1403.
• 34. Bellemans, T., et al., Implementation framework and development trajectory of FEATHERS activity-based simulation platform. Transportation Research Record: Journal of the Transportation Research Board, 2010(2175): p. 111-119.
• 35. Balmer, M., K. Axhausen, and K. Nagel, Agent-based demand-modeling framework for large-scale microsimulations. Transportation Research Record: Journal of the Transportation Research Board, 2006(1985): p. 125-134.
• 36. Nagel, K., R.L. Beckman, and C.L. Barrett. TRANSIMS for transportation planning . in In 6th Int. Conf. on Computers in Urban Planning and Urban Management . 1999. Citeseer.
• 37. Adnan, M., et al. Simmobility: A multi-scale integrated agent-based simulation platform . in 95th Annual Meeting of the Transportation Research Board Forthcoming in Transportation Research Record . 2016.
• 38. Auld, J., et al., POLARIS: Agent-based modeling framework development and implementation for integrated travel demand and network and operations simulations. Transportation Research Part C: Emerging Technologies, 2016. 64 : p. 101-116.
• 39. Kagho, G.O., M. Balac, and K.W. Axhausen, Agent-based models in transport planning: Current state, issues, and expectations. Procedia Computer Science, 2020. 170 : p. 726-732.
• 40. Miller, E.J., Travel demand models, the next generation: Boldly going where no-one has gone before , in Mapping the Travel Behavior Genome . 2020, Elsevier. p. 29-46.
• 41. Nurul Habib, K., E. Miller, and K. Axhausen, Weekly rhythm in joint time expenditure for all at-home and out-of-home activities: application of Kuhn-Tucker demand system model using multiweek travel diary data. Transportation Research Record: Journal of the Transportation Research Board, 2008(2054): p. 64-73.
• 42. Anda, C., A. Erath, and P.J. Fourie, Transport modelling in the age of big data. International Journal of Urban Sciences, 2017. 21 (sup1): p. 19-42.
• 43. Zhao, Z., J. Zhao, and H.N. Koutsopoulos, Individual-Level Trip Detection using Sparse Call Detail Record Data based on Supervised Statistical Learning . 2016.
• 44. Caceres, N., J. Wideberg, and F. Benitez, Deriving origin–destination data from a mobile phone network. IET Intelligent Transport Systems, 2007. 1 (1): p. 15-26.
• 45. Wang, P., et al., Understanding road usage patterns in urban areas. Scientific reports, 2012. 2 : p. 1001.
• 46. Farrahi, K. and D. Gatica-Perez, Discovering routines from large-scale human locations using probabilistic topic models. ACM Transactions on Intelligent Systems and Technology (TIST), 2011. 2 (1): p. 3.
• 47. Phithakkitnukoon, S., et al. Activity-aware map: Identifying human daily activity pattern using mobile phone data . in International Workshop on Human Behavior Understanding . 2010. Springer.
• 48. Widhalm, P., et al., Discovering urban activity patterns in cell phone data. Transportation, 2015. 42 (4): p. 597-623.
• 49. Yin, M., et al., A generative model of urban activities from cellular data. IEEE Transactions on Intelligent Transportation Systems, 2017.
• 50. Schneider, C.M., et al., Unravelling daily human mobility motifs. Journal of The Royal Society Interface, 2013. 10 (84): p. 20130246.
• 51. Jiang, S., et al. A review of urban computing for mobile phone traces: current methods, challenges and opportunities . in Proceedings of the 2nd ACM SIGKDD international workshop on Urban Computing . 2013. ACM.
• 52. Jiang, S., J. Ferreira, and M.C. Gonzalez, Activity-based human mobility patterns inferred from mobile phone data: A case study of Singapore. IEEE Transactions on Big Data, 2017. 3 (2): p. 208-219.
• 53. Zilske, M. and K. Nagel, Studying the accuracy of demand generation from mobile phone trajectories with synthetic data. Procedia Computer Science, 2014. 32 : p. 802-807.
• 54. Zilske, M. and K. Nagel, A simulation-based approach for constructing all-day travel chains from mobile phone data. Procedia Computer Science, 2015. 52 : p. 468-475.
• 55. Bassolas, A., et al., Mobile phone records to feed activity-based travel demand models: MATSim for studying a cordon toll policy in Barcelona. Transportation Research Part A: Policy and Practice, 2019. 121 : p. 56-74.
• 56. Devillaine, F., M. Munizaga, and M. Trépanier, Detection of activities of public transport users by analyzing smart card data. Transportation Research Record: Journal of the Transportation Research Board, 2012(2276): p. 48-55.
• 57. Ceapa, I., C. Smith, and L. Capra. Avoiding the crowds: understanding tube station congestion patterns from trip data . in Proceedings of the ACM SIGKDD international workshop on urban computing . 2012. ACM.
• 58. da Silva, T.L.C., J.A. de Macêdo, and M.A. Casanova. Discovering frequent mobility patterns on moving object data . in Proceedings of the Third ACM SIGSPATIAL International Workshop on Mobile Geographic Information Systems . 2014. ACM.
• 59. Liu, Y., et al., Intelligent bus routing with heterogeneous human mobility patterns. Knowledge and Information Systems, 2017. 50 (2): p. 383-415.
• 60. Zhang, F., et al., Spatiotemporal segmentation of metro trips using smart card data. IEEE Transactions on Vehicular Technology, 2016. 65 (3): p. 1137-1149.
• 61. Zhao, J., et al., Estimation of passenger route choice pattern using smart card data for complex metro systems. IEEE Transactions on Intelligent Transportation Systems, 2017. 18 (4): p. 790-801.
• 62. Bouman, P., et al., Recognizing demand patterns from smart card data for agent-based micro-simulation of public transport. 2012.
• 63. Chakirov, A. and A. Erath, Activity identification and primary location modelling based on smart card payment data for public transport. [Working paper/Transport and Spatial Planning], 2012. 786 .
• 64. Fourie, P.J., et al., Using smartcard data for agent-based transport simulation , in Public Transport Planning with Smart Card Data . 2016, CRC Press. p. 133-160.
• 65. Zhu, Y., Extract the Spatiotemporal Distribution of Transit Trips from Smart Card Transaction Data: A Comparison Between Shanghai and Singapore , in Big Data Support of Urban Planning and Management . 2018, Springer. p. 297-315.
• 66. Patterson, D.J., et al. Inferring high-level behavior from low-level sensors . in International Conference on Ubiquitous Computing . 2003. Springer.
• 67. Schuessler, N. and K. Axhausen, Processing raw data from global positioning systems without additional information. Transportation Research Record: Journal of the Transportation Research Board, 2009(2105): p. 28-36.
• 68. Rudloff, C. and M. Ray, Detecting travel modes and profiling commuter habits solely based on GPS data . 2010.
• 69. McGowen, P. and M. McNally. Evaluating the potential to predict activity types from GPS and GIS data . in Transportation Research Board 86th Annual Meeting, Washington . 2007. Citeseer.
• 70. Demissie, M.G., et al., Inferring passenger travel demand to improve urban mobility in developing countries using cell phone data: a case study of Senegal. IEEE Transactions on Intelligent Transportation Systems, 2016. 17 (9): p. 2466-2478.
• 71. Ndibatya, I., J. Coetzee, and T. Booysen, Mapping the informal public transport network in Kampala with smartphones: international. Civil Engineering = Siviele Ingenieurswese, 2017. 2017 (v25i1): p. 35-40.
• 72. Saddier, S., et al., Mapping the Jitney network with smartphones in Accra, Ghana: the AccraMobile experiment. Transportation Research Record: Journal of the Transportation Research Board, 2016(2581): p. 113-122.
• 73. Bisseling, R., et al., Inferring transportation modes from smartphone sensors. Proceedings 106th European Study Group Mathematics with Industry, 2016: p. 21-34.
• 74. Cheng, Z., et al., Exploring millions of footprints in location sharing services. ICWSM, 2011. 2011 : p. 81-88.
• 75. Rashidi, T.H., et al., Exploring the capacity of social media data for modelling travel behaviour: Opportunities and challenges. Transportation Research Part C: Emerging Technologies, 2017. 75 : p. 197-211.
• 76. Golder, S.A. and M.W. Macy, Digital footprints: Opportunities and challenges for online social research. Annual Review of Sociology, 2014. 40 .
• 77. Cramer, H., M. Rost, and L.E. Holmquist. Performing a check-in: emerging practices, norms and'conflicts' in location-sharing using foursquare . in Proceedings of the 13th international conference on human computer interaction with mobile devices and services . 2011. ACM.
• 78. Maghrebi, M., et al. Complementing travel diary surveys with Twitter data: application of text mining techniques on activity location, type and time . in Intelligent Transportation Systems (ITSC), 2015 IEEE 18th International Conference on . 2015. IEEE.
• 79. Zhu, Z., U. Blanke, and G. Tröster. Inferring travel purpose from crowd-augmented human mobility data . in Proceedings of the First International Conference on IoT in Urban Space . 2014. ICST (Institute for Computer Sciences, Social-Informatics and Telecommunications Engineering).
• 80. Hasan, S. and S.V. Ukkusuri, Social contagion process in informal warning networks to understand evacuation timing behavior. Journal of Public Health Management and Practice, 2013. 19 : p. S68-S69.
• 81. Noulas, A., et al., A tale of many cities: universal patterns in human urban mobility. PloS one, 2012. 7 (5): p. e37027.
• 82. Jurdak, R., et al., Understanding human mobility from Twitter. PloS one, 2015. 10 (7): p. e0131469.
• 83. Cebelak, M.K., Location-based social networking data: doubly-constrained gravity model origin-destination estimation of the urban travel demand for Austin, TX. 2013.
• 84. Lee, J.H., et al., Activity space estimation with longitudinal observations of social media data. Transportation, 2016. 43 (6): p. 955-977.
• 85. Lee, J.H., S. Gao, and K.G. Goulias. Can Twitter data be used to validate travel demand models . in 14th International Conference on Travel Behaviour Research . 2015.
• 86. Hasnat, M.M., et al., Destination choice modeling using location-based social media data. Journal of choice modelling, 2019. 31 : p. 22-34.
• 87. Lian, D. and X. Xie. Collaborative activity recognition via check-in history . in Proceedings of the 3rd ACM SIGSPATIAL International Workshop on Location-Based Social Networks . 2011. ACM.
• 88. Pianese, F., et al. Discovering and predicting user routines by differential analysis of social network traces . in World of Wireless, Mobile and Multimedia Networks (WoWMoM), 2013 IEEE 14th International Symposium and Workshops on a . 2013. IEEE.
• 89. Coffey, C. and A. Pozdnoukhov. Temporal decomposition and semantic enrichment of mobility flows . in Proceedings of the 6th ACM SIGSPATIAL International Workshop on Location-Based Social Networks . 2013. ACM.
• 90. Hasan, S. and S.V. Ukkusuri, Location contexts of user check-ins to model urban geo life-style patterns. PloS one, 2015. 10 (5): p. e0124819.
• 91. Medina, S.A.O., Inferring weekly primary activity patterns using public transport smart card data and a household travel survey. Travel Behaviour and Society, 2016.
• 92. Zhang, A., et al. Multi-day activity-travel pattern sampling based on single-day data . in 97th Annual Meeting of the Transportation Research Board (TRB 2018) . 2018. TRB Annual Meeting.
• 93. Horni, A. and K.W. Axhausen, MATSim Agent Heterogeneity and a One-Week Scenario. ETH, Eidgenössische Technische Hochschule Zürich, IVT, Institut für Verkehrsplanung und Transportsysteme, 2012.
• 94. Ordóñez Medina, S.A., A. Erath, and K.W. Axhausen. Simulating Urban Transport for a Week Time Horizon Using MATSim . in 3rd Workshop on Time Use Observatory (TUO 3) . 2012. Complex Engineering Systems Institute (ISCI).
• 95. Ordóñez Medina, S.A. Recognizing personalized flexible activity patterns . in 14th International Conference on Travel Behavior Research (IATBR 2015) . 2015. IVT, ETH Zurich.
• 96. Lee, J.H., Travel Behavior Dynamics in Space and Time . 2016, University of California, Santa Barbara.
• 97. Balac, M., M. Janzen, and K.W. Axhausen. Alternative Approach to Scoring in MATSim and how it affects Activity Rescheduling . in 97th Annual Meeting of the Transportation Research Board (TRB 2018) . 2018. TRB Annual Meeting.
• 98. Koppelman, F.S. and C.G. Wilmot, Transferability analysis of disaggregate choice models. Transportation Research Record, 1982. 895 : p. 18-24.
• 99. Yasmin, F., C. Morency, and M.J. Roorda, Assessment of spatial transferability of an activity-based model, TASHA. Transportation Research Part A: Policy and Practice, 2015. 78 : p. 200-213.
• 100. Sikder, S., et al., Spatial transferability of travel forecasting models: a review and synthesis. International Journal of Advances in Engineering Sciences and Applied Mathematics, 2013. 5 (2-3): p. 104-128.
• 101. Arentze, T., et al., Spatial transferability of the Albatross model system: Empirical evidence from two case studies. Transportation Research Record: Journal of the Transportation Research Board, 2002(1805): p. 1-7.
• 102. Inc, P.C., Lake Tahoe Resident and Visitor Model: Model Description and Final Results. 2007.
• 103. Bowman, J.L., et al. Making advanced travel forecasting models affordable through model transferability . in 93rd Transportation Research Board Annual Meeting, Washington DC, USA . 2014.
• 104. Goulias, K.G., et al. Simulator of activities, greenhouse emissions, networks, and travel (SimAGENT) in Southern California . in 91st annual meeting of the Transportation Research Board, Washington, DC . 2012.
• 105. Sivakumar, A., S. Le Vine, and J. Polak. An activity-based travel demand model for London . in European Transport Conference, 2010Association for European Transport . 2010.
• 106. Yasmin, F., C. Morency, and M.J. Roorda, Macro-, meso-, and micro-level validation of an activity-based travel demand model. Transportmetrica A: Transport Science, 2017. 13 (3): p. 222-249.
• 107. Ziemke, D., K. Nagel, and C. Bhat, Integrating CEMDAP and MATSim to increase the transferability of transport demand models. Transportation Research Record: Journal of the Transportation Research Board, 2015(2493): p. 117-125.
• 108. Bowman, J.L. and M. Bradley, Testing Spatial Transferability of Activity-Based Travel Forecasting Models. Transportation Research Record: Journal of the Transportation Research Board, 2017(2669): p. 62-71.
• 109. Linha, H.T., et al., Exploring the Transferability of FEATHERS–An Activity Based Travel Demand Model–For Ho Chi Minh City, Vietnam. Journal of Traffic and Transportation Management, 2019. 1 (2): p. 01-09.
• 110. Shiftan, Y., The advantage of activity-based modelling for air-quality purposes: theory vs practice and future needs. Innovation: The European Journal of Social Science Research, 2000. 13 (1): p. 95-110.
• 111. Shiftan, Y. and J. Suhrbier, The analysis of travel and emission impacts of travel demand management strategies using activity-based models. Transportation, 2002. 29 (2): p. 145-168.
• 112. Beckx, C., et al., An integrated activity-based modelling framework to assess vehicle emissions: approach and application. Environment and Planning B: Planning and Design, 2009. 36 (6): p. 1086-1102.
• 113. Dons, E., et al., Using an activity-based framework to determine effects of a policy measure on population exposure to nitrogen dioxide. Transportation Research Record: Journal of the Transportation Research Board, 2011(2233): p. 72-79.
• 114. Pebesma, E., et al. Uncertainty in exposure to air pollution . in EGU General Assembly Conference Abstracts . 2013.
• 115. Hatzopoulou, M., E. Miller, and B. Santos, Integrating vehicle emission modeling with activity-based travel demand modeling: case study of the Greater Toronto, Canada, Area. Transportation Research Record: Journal of the Transportation Research Board, 2007(2011): p. 29-39.
• 116. Hao, J., M. Hatzopoulou, and E. Miller, Integrating an activity-based travel demand model with dynamic traffic assignment and emission models: Implementation in the Greater Toronto, Canada, area. Transportation Research Record: Journal of the Transportation Research Board, 2010(2176): p. 1-13.
• 117. Hatzopoulou, M., J.Y. Hao, and E.J. Miller, Simulating the impacts of household travel on greenhouse gas emissions, urban air quality, and population exposure. Transportation, 2011. 38 (6): p. 871.
• 118. Vallamsundar, S., et al., Maternal Exposure to Traffic-Related Air Pollution Across Different Microenvironments. Journal of Transport & Health, 2016. 3 (2): p. S72.
• 119. Shabanpour, R., et al., Investigating the applicability of ADAPTS activity-based model in air quality analysis. Travel Behaviour and Society, 2017.
• 120. Wu, Y. and G. Song, The Impact of Activity-Based Mobility Pattern on Assessing Fine-Grained Traffic-Induced Air Pollution Exposure. International journal of environmental research and public health, 2019. 16 (18): p. 3291.
• 121. Shliselberg, R. and M. Givoni, Motility as a policy objective. Transport reviews, 2018. 38 (3): p. 279-297.
• 122. Abou-Zeid, M. and M. Ben-Akiva, Well-being and activity-based models. Transportation, 2012. 39 (6): p. 1189-1207.
• 123. Stanley, J.K., et al., Mobility, social exclusion and well-being: Exploring the links. Transportation research part A: policy and practice, 2011. 45 (8): p. 789-801.
• 124. Hensher, D.A., Measurement of the valuation of travel time savings. Journal of Transport Economics and Policy (JTEP), 2001. 35 (1): p. 71-98.
• 125. Metz, D., The myth of travel time saving. Transport reviews, 2008. 28 (3): p. 321-336.
• 126. El-Geneidy, A.M. and D.M. Levinson, Access to destinations: Development of accessibility measures. 2006.
• 127. de Lima, I.V., et al., Dynamic Modeling of Activity Happiness: An Investigation of the Intra-activity Hedonic Treadmill , in Quality of Life and Daily Travel . 2018, Springer. p. 95-118.
• 128. Guo, W., Y. Chai, and M.-P. Kwan, Travel-related exposure to air pollution and its socio-environmental inequalities: Evidence from a week-long GPS-based travel diary dataset , in Spatiotemporal Analysis of Air Pollution and Its Application in Public Health . 2020, Elsevier. p. 293-309.
• 129. Castiglione, J., et al. Building an Integrated Activity-Based and Dynamic Network Assignment Model . in 3rd Transportation Research Board Conference on Innovations in Travel Modeling, Tempe, Ariz . 2010.
• 130. Bekhor, S., C. Dobler, and K. Axhausen, Integration of activity-based and agent-based models: case of Tel Aviv, Israel. Transportation Research Record: Journal of the Transportation Research Board, 2011(2255): p. 38-47.
• 131. Lin, D.-Y., et al., Evacuation planning using the integrated system of activity-based modeling and dynamic traffic assignment. Transportation Research Record: Journal of the Transportation Research Board, 2009(2132): p. 69-77.
• 132. Liu, P., et al., Dynamic activity-travel assignment in multi-state supernetworks. Transportation Research Part B: Methodological, 2015. 81 : p. 656-671.
• 133. Pendyala, R.M., et al., Integrated Land Use–Transport Model System with Dynamic Time-Dependent Activity–Travel Microsimulation. Transportation Research Record, 2012. 2303 (1): p. 19-27.
• 134. Pendyala, R.M., et al., Paradigms for integrated modeling of activity-travel demand and network dynamics in an era of dynamic mobility management . 2017.
• 135. Zockaie, A., et al., Activity-Based Model with Dynamic Traffic Assignment and Consideration of Heterogeneous User Preferences and Reliability Valuation: Application to Toll Revenue Forecasting in Chicago, Illinois. Transportation Research Record: Journal of the Transportation Research Board, 2015(2493): p. 78-87.
• 136. Rieser, M., et al., Agent-oriented coupling of activity-based demand generation with multiagent traffic simulation. Transportation Research Record, 2007. 2021 (1): p. 10-17.
• 137. Hao, J.Y., M. Hatzopoulou, and E.J. Miller, Integrating an activity-based travel demand model with dynamic traffic assignment and emission models: Implementation in the Greater Toronto, Canada, area. Transportation Research Record, 2010. 2176 (1): p. 1-13.
• 138. Lin, D.-Y., et al., Integration of activity-based modeling and dynamic traffic assignment. Transportation Research Record, 2008. 2076 (1): p. 52-61.
• 139. Hadi, M., et al., Dynamic, Integrated Model System: Jacksonville-Area Application . 2014.
• 140. Javanmardi, M., J. Auld, and K. Mohammadian. Integration of TRANSIMS with the ADAPTS Activity-based Model . in Fourth TRB Conference on Innovations in Travel Modeling (ITM), Tampa, FL . 2011.
• 141. Zockaie, A., et al., Activity-based model with dynamic traffic assignment and consideration of heterogeneous user preferences and reliability valuation: application to toll revenue forecasting in Chicago, Illinois. Transportation Research Record, 2015. 2493 (1): p. 78-87.
• 142. Vovsha, P., et al. Integrated model of travel demand and network simulation . in Proceedings of the 6th Conference on Innovations in Travel Modeling (ITM), TRB, Denver, CO . 2016.
• 143. Cipriani, E., et al., Integration between activity-based demand models and multimodal assignment: some empirical evidences. Case Studies on Transport Policy, 2020.
• 144. Adnan, M., et al., Integrated agent-based microsimulation framework for examining impacts of mobility-oriented policies. Personal and Ubiquitous Computing, 2020: p. 1-13.
• 145. Baghestani, A., et al., Evaluating the Traffic and Emissions Impacts of Congestion Pricing in New York City. Sustainability, 2020. 12 (9): p. 3655.
• 146. Cepolina, E.M. and A. Farina, A new shared vehicle system for urban areas. Transportation Research Part C: Emerging Technologies, 2012. 21 (1): p. 230-243.
• 147. Ciari, F., M. Balac, and K.W. Axhausen, Modeling Carsharing with the Agent-Based Simulation MATSim: State of the Art, Applications, and Future Developments. Transportation Research Record: Journal of the Transportation Research Board, 2016(2564): p. 14-20.
• 148. Balac, M., F. Ciari, and K.W. Axhausen, Modeling the impact of parking price policy on free-floating carsharing: Case study for Zurich, Switzerland. Transportation Research Part C: Emerging Technologies, 2017. 77 : p. 207-225.
• 149. Heilig, M., et al., Implementation of free-floating and station-based carsharing in an agent-based travel demand model. Travel Behaviour and Society, 2018. 12 : p. 151-158.
• 150. Li, Q., et al., Incorporating free-floating car-sharing into an activity-based dynamic user equilibrium model: A demand-side model. Transportation Research Part B: Methodological, 2018. 107 : p. 102-123.
• 151. Choo, S., P.L. Mokhtarian, and I. Salomon, Does telecommuting reduce vehicle-miles traveled? An aggregate time series analysis for the US. Transportation, 2005. 32 (1): p. 37-64.
• 152. Kim, S.-N., Is telecommuting sustainable? An alternative approach to estimating the impact of home-based telecommuting on household travel. International Journal of Sustainable Transportation, 2017. 11 (2): p. 72-85.
• 153. Zhu, P. and S.G. Mason, The impact of telecommuting on personal vehicle usage and environmental sustainability. International Journal of Environmental Science and Technology, 2014. 11 (8): p. 2185-2200.
• 154. Paleti, R. and I. Vukovic, Telecommuting and Its Impact on Activity–Time Use Patterns of Dual-Earner Households. Transportation Research Record: Journal of the Transportation Research Board, 2017(2658): p. 17-25.
• 155. Helminen, V. and M. Ristimäki, Relationships between commuting distance, frequency and telework in Finland. Journal of Transport Geography, 2007. 15 (5): p. 331-342.
• 156. Pouri, Y. and C. Bhat, On Modeling Choice and Frequency of Home-Based Telecommunting. Transportation Research Record: Journal of the Transportation Research Board, 2003(1858): p. 55-60.
• 157. Sener, I.N. and C.R. Bhat, A copula-based sample selection model of telecommuting choice and frequency. Environment and Planning A, 2011. 43 (1): p. 126-145.
• 158. Shabanpour, R., et al., Analysis of telecommuting behavior and impacts on travel demand and the environment. Transportation Research Part D: Transport and Environment, 2018. 62 : p. 563-576.
• 159. Pirdavani, A., et al., Traffic Safety Implications of Travel Demand Management Policies: The Cases of Teleworking and Fuel Cost Increase , in Transportation Systems and Engineering: Concepts, Methodologies, Tools, and Applications . 2015, IGI Global. p. 1082-1107.
• 160. Zhang, L., et al., Integrating an agent-based travel behavior model with large-scale microscopic traffic simulation for corridor-level and subarea transportation operations and planning applications. Journal of Urban Planning and Development, 2013. 139 (2): p. 94-103.
• 161. Yang, D., INTEGRATING ACTIVITY-BASED TRAVEL DEMAND AND DYNAMIC TRAFFIC ASSIGNMENT MODEL: A BEHAVIORAL USER EQUILIBRIUM APPROACH . 2018.
• 162. Frank, L.D., et al., Many pathways from land use to health: associations between neighborhood walkability and active transportation, body mass index, and air quality. Journal of the American planning Association, 2006. 72 (1): p. 75-87.

© 2020 The Author(s). Licensee IntechOpen. This chapter is distributed under the terms of the Creative Commons Attribution 3.0 License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Continue reading from the same book

Edited by Stefano de Luca

Published: 28 July 2021

By Joan Harvey

444 downloads

By Carlos Hugo Criado del Valle and Parichehr Scharif...

394 downloads

By Francesco Viti, Marco Rinaldi and Georgios Laskari...

553 downloads

Activity-Based Travel Demand Models: A Primer (2014)

Chapter: 5 case examples.

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

131 SHRP 2 C10A JACKSONVILLE, FLORIDA, AND BURLINGTON, VERMONT Objective The primary objective of the SHRP 2 C10A project was to make operational a regional-scale dynamic integrated model and to demonstrate the model’s performance through validation tests and policy analyses. The model system was designed to cap- ture changes in demand, such as time-of-day choice and peak-spreading, destination, and mode and route choice, in response to capacity and operational improvements such as signal coordination, freeway management, and variable tolls. An additional goal was to develop a model system that could be transferred to other regions, as well to incorporate fi ndings from other SHRP 2 efforts. The model system was imple- mented in two regions: Jacksonville, Florida, and Burlington, Vermont. In both re- gions, implementation of the model system was primarily performed by a consultant team, with the Jacksonville MPO and the Florida DOT providing data and support (Resource Systems Group et al. 2014). The information presented in this example was gathered from project reports and interviews with project team members. Model System Design and Components The SHRP 2 C10A model system comprises two primary components. They are DaySim and the TRANSIMS Router and Microsimulator. DaySim is an activity-based travel demand forecast model that predicts household and person travel choices at a parcel level on a minute-by-minute basis. The TRANSIMS Router and Microsimulator is dynamic network assignment and network simulation software that track vehicles on a second-by-second basis. DaySim simulates 24-hour itineraries for individuals with spatial resolution as fi ne as individual parcels and temporal resolution as fi ne as single minutes, so it can generate outputs at the level of resolution required as input to dynamic traffi c simulation. The TRANSIMS network microsimulation process assigns a sequence of trips or tours for individual household persons between specifi c 5 CASE EXAMPLES

132 Part 2: ISSUES IN ADOPTING INTEGRATED DYNAMIC MODELS SYSTEMS activity locations to paths on a second-by-second basis for a full travel day. The net- work includes detailed information regarding the operational characteristics of the transportation facilities that may vary by time of day and by vehicle or traveler type such as the number of lanes; the lane use restrictions; traffic controls, signal timing, and phasing plans; turning restrictions; and tolls and parking fees (Resource Systems Group et al. 2014). The C10A model system was implemented for two regions. The Jacksonville model includes four counties in northern Florida with a population of approximately 1.2 million people, while the Burlington model includes one county in Vermont with a population of approximately 150,000 people. The model system employs multiple spatial resolutions—the base spatial data describing employment and households are at the level of individual parcels, while the network performance indicators are avail- able at either the TAZ level or the activity location (AL) level. The model system also employs multiple temporal resolutions. On the demand model side, the core time- of-day models within DaySim operate at temporal resolutions as fine as 10 minutes and are subsequently disaggregated to individual minutes. On the supply model side, TRANSIMS follows vehicles on a second-by-second basis, measures of link-level net- work performance are typically collected using 5-minute intervals, and measures of O-D network performance (or skims) are also generated using temporal resolutions as fine as 10 minutes, consistent with the demand side time-of-day models. The DaySim model incorporates significant typological detail, including basic persontypes, as well as identifying trip-specific values of time reflective of travel purpose, traveler income, and mode. This value-of-time information is incorporated into the TRANSIMS net- work assignment process using approximately 50 value-of-time classes. TRANSIMS also generates network skims by value-of-time class for input into DaySim. The model system is configured to run a fixed number of assignment iterations and system itera- tions and is designed to achieve sufficient levels of convergence as necessary to gener- ate meaningful performance metrics for planning purpose (Resource Systems Group et al. 2014). Lessons Learned Developing the inputs to the DaySim activity-based demand model components was relatively straightforward, though significant cleaning was required. Transferring the DaySim activity-based demand component from Sacramento to Jacksonville radically reduced the amount of time required to implement the activity-based demand model component of the model system. In contrast, developing detailed and usable networks for microsimulation required a significant level of effort, although this effort was miti- gated by using TRANSIMS tools to perform network development tasks and the avail- ability of spatially detailed network data. Correcting topological errors; resolving attri- bute discontinuities; coding intersection controls; and iteratively evaluating, adjusting, and testing the networks by running simulations is time-consuming. In addition, there are numerous challenges when developing future-year or alternative network scenarios (Resource Systems Group et al. 2014).

133 Chapter 5: CASE EXAMPLES As noted in the SHRP 2 C10A final report, “Configuring DaySim to generate temporally, spatially, and behaviorally detailed travel demand information for use in TRANSIMS was straightforward, as was configuring TRANSIMS to generate the skims for input to DaySim. More sophisticated methods of providing TRANSIMS- based impedances to DaySim, such as implementing efficient multistage sampling of destinations (and corresponding impedances) at strategic points in the DaySim looping process or integrating DaySim and TRANSIMS so that DaySim can call TRANSIMS to extract the required measures quickly, could potentially be implemented.” (Resource Systems Group et al. 2014). The new model system is more sensitive to a wider range of policies than a tradi- tional travel demand model system, and this sensitivity is further enhanced by the detailed representation of temporal dimension. Extensive testing of the model system was necessary to determine the number of network assignment and model system iter- ations required to ensure that differences between alternative scenario model results were attributable to these policy and investments and not obscured by noise in the model system. Extracting, managing, and interpreting these results was not difficult; however, the level of effort required to effectively test different types of improvements varied widely, from as little as an hour to as more than a week. It is safe to say that a higher degree of knowledge and patience is required when interacting with the new integrated model system than is required when using a traditional trip-based model system (Resource Systems Group et al. 2014). SHRP 2 C10B SACRAMENTO, CALIFORNIA Objective As with the SHRP 2 C10A project, the primary objective of the SHRP 2 C10B project was to make operational a regional-scale dynamic integrated model and to demon- strate the model’s performance through validation tests and policy analyses. However, there are two notable distinctions between the scopes of the C10B and C10A projects. First, the size of the Sacramento, California, region used in the C10B project is approx- imately twice as big as the Jacksonville region used in the C10A project, and more than 10 times as large as the Burlington region used in the C10A project. Second, and more significantly, the C10B project included a dynamic transit demand network assignment model in addition to a dynamic roadway network assignment model. Even though many dynamic roadway network assignment models represent interactions between transit vehicles and private vehicles, there are very few models that provide the capa- bility to assign transit demand and represent the effect of this demand on network per- formance. Although the development of the integrated model system was primarily led by a consultant team, SACOG staff were actively involved in the C10B effort, such as performing the model sensitivity test runs (T. Rossi, personal communication, Oct. 17, 2013). The information presented in this example was gathered primarily from inter- views with project team members and from project reports.

134 Part 2: ISSUES IN ADOPTING INTEGRATED DYNAMIC MODELS SYSTEMS Model System Design and Components The SHRP 2 C10B model system comprises three primary components: DaySim, Dynus-T, and FAST-TrIPs. DaySim is a travel demand forecast model that predicts household and person travel choices at a parcel level on a minute-by-minute basis. Dynus-T is the dynamic roadway traffic assignment tool, which tracks vehicles on the network on a second-by-second basis, and FAST-TrIPs is the dynamic transit demand assignment tool, which tracks transit travelers on a second-by-second basis ( Cambridge Systematics, Inc. et al. 2014). DaySim simulates 24-hour itineraries for individuals with spatial resolution as fine as individual parcels and temporal reso- lution as fine as single minutes, so it can generate outputs at the level of resolution required as input to dynamic traffic simulation. Dynus-T assigns a sequence of trips or tours for individual household persons between specific activity locations to paths on a second-by-second basis for a full travel day and incorporates significant capabili- ties to adjust the input demand in order to generate more realistic results. Dynus-T operates at a mesoscopic scale, which is different from the microscopic simulations of the TRANSIMS, TransModeler, and Dynameq software used in the other integrated model development efforts described in this document, although Dynus-T shares some similarities with microscopic car-following-based models (Cambridge Systematics, Inc. et al. 2014). FAST-TrIPs is a transit assignment tool that is designed to accurately represent transit operations, to capture the operational dynamics of transit vehicles, to provide both schedule-based and frequency-based transit traveler assignment, and to generate skims for feedback to the activity-based travel demand model (Cambridge Systematics, Inc. et al. 2014). The C10B model system was implemented in the Sacramento, California, region, which includes approximately 2.3 million people. Like the C10A project, the C10B model system employs multiple spatial resolutions—the base spatial data describing employment and households are at the level of individual parcels, while the network performance indicators are available at the TAZ level. The model system also employs multiple temporal resolutions. On the demand model side, the core time-of-day models within DaySim operate at temporal resolutions of 30 minutes, which are sub- sequently disaggregated to individual minutes. On the supply model side, Dynus-T and FAST-TrIPs follow vehicles on a second-by-second basis, and ultimately skims of O-D network performance are generated using a temporal resolution of 30 minutes, consis- tent with the demand side time-of-day models. The model system is configured to run a fixed number of assignment iterations and system iterations and is designed to achieve sufficient levels of convergence as necessary to generate meaningful performance met- rics for planning purposes (T. Rossi, personal communication, Oct. 17, 2013). Lessons Learned The most important lesson learned from this effort is that it demonstrates the feasi- bility of implementing a regional-scale integrated activity-based model and dynamic traffic and transit assignment models. The project clearly illustrated the potential ben- efits of a more continuous representation of time such as the ability to generate more

135 Chapter 5: CASE EXAMPLES detailed network performance skims, as well as the ability to more precisely character- ize the location, extent, and duration of congestion (T. Rossi, personal communica- tion, Oct. 17, 2013). However, implementing the model system was a significant undertaking. Exten- sive efforts were required to develop and calibrate the roadway and transit assign- ment models, and additional efforts are likely required to achieve a level of confidence required to support project evaluations. Applying the model was also complicated by the relatively long model system run times, and by the fact that the integrated model system requires a relatively high level of modeler involvement to execute a complete integrated run. Interpretation of model results also proved to be challenging, and fur- ther work is required in order to ensure that the network assignment models and the overall model system are reasonably well converged before being suitable to sup- port policy and investment analyses. Stochasticity in the model results appears to be an issue that will require further investigation. Finally, the project team felt that the iterative development and expansion of modeled area may not be the most effective method for getting to a full regional model implementation (B. Griesenbeck, personal communication, Oct. 17, 2013). SAN FRANCISCO COUNTY TRANSPORTATION AUTHORITY’S “DTA ANYWAY” Objective The goal of the San Francisco County Transportation Authority’s “DTA Anyway” project was to develop an application-ready tool that the SFCTA could use to evalu- ate projects throughout the city. SFCTA staff were particularly interested in under- standing the effects of congestion pricing on transit performance and traffic diversion, representing operational strategies, and producing realistic traffic flows in which the forecast demand does not exceed assumed capacities. This citywide dynamic network model built on an earlier dynamic traffic (or network) assignment (DTA) model built for the northwest quadrant of the city that had been successfully applied to analyze construction phasing for a major roadway project. Beyond being able to use the model to support SFCTA’s planning activities, SFCTA staff also sought to assist future DTA deployment efforts by building a toolkit in Python programming language that pro- vides capabilities such as network data exchange and conversion procedures and reporting capabilities. Additionally, SFCTA staff tried to fully document the process and assumptions used in implementing the citywide DTA model and to reveal DTA performance in the context of a congested grid network in which there is significant interaction with transit vehicles and demand (Parsons Brinckerhoff and San Francisco County Transportation Authority 2012). Model System Design and Components The SFCTA DTA Anyway model system comprises two primary components: the SF-CHAMP activity-based model system and the Dynameq network simulation model. SF-CHAMP is an activity-based travel demand forecast model that predicts household and person travel choices. Dynameq is microscopic traffic simulation model

136 Part 2: ISSUES IN ADOPTING INTEGRATED DYNAMIC MODELS SYSTEMS with sufficient detail to consider lane-level modeling. The SF-CHAMP activity-based model generates demand for the entire San Francisco Bay Area region, while the DTA Anyway model implementation covers San Francisco County only. In order to bridge these different spatial extents, a subarea extraction process using a static network assignment model is used in which vehicle flows into and out of San Francisco are sum- marized. This method can be effectively applied in San Francisco because of the unique geographical features that define the city. SF-CHAMP forecasts DAPs for all regional residents using a spatial resolution of very small TAZs within San Francisco and a temporal resolution of multihour time periods. SF-CHAMP is sensitive to the impact of travel times and costs on time of day, mode, destination, and activity generation, and generates lists of person-level tours and trips. However, these tours and trips are not used directly in the model. Rather the trips lists are aggregated to matrices of flows by time of day and mode, these flows are assigned to SF-CHAMP’s static model networks, and subarea matrices are derived from this assignment. The Dynameq traffic microsimulation model assigns discrete vehicle trips to a detailed San Francisco network. This network includes information on the actual sig- nal and timing plans for all traffic signals in the city (there are more than 1,100) as well as the locations of other intersection controls, such as more than 3,000 stop-sign locations. The network also includes detailed information regarding the operational characteristics of the transportation facilities that may vary by time of day and by vehicle type or traveler type such as the number of lanes, lane use restrictions, traffic controls and signal timing and phasing plans, turning restrictions, tolls, and parking fees. The network simulation also includes transit vehicles, which are an important segment of the vehicle fleet operating in transit-rich San Francisco. A key challenge in integrating the SF-CHAMP activity-based demand model and the Dynameq traffic microsimulation model is the different temporal resolutions used by these two tools. It is necessary to temporally disaggregate the broad time-period demand produced by the activity-based model down to the finer time slices used by Dynameq. In the San Francisco Dynameq network model, changes in network performance by time of day that are used to build paths are represented using 7.5-minute intervals, although the simulation of vehicle interactions uses a significantly finer temporal resolution. The network simulation is performed only for the 3-hour p.m. peak period, although a one-hour warm-up period, and a one-hour cool-down period are also simulated. During this time period, approximately 450,000 vehicle trips are assigned. It should be noted that because of the limited temporal and spatial extents of the traffic micro- simulation model, it is not feasible to generate skims for feeding back input to the activity-based model. Thus, the integration of the model is one way. Lessons Learned SFCTA staff and their consultant team members learned a number of meaningful les- sons from the development of the dynamic traffic model for the city and the inte- gration of activity-based demand into this model. From a practical perspective, the automated procedures for aligning the Dynameq and SF-CHAMP data assumptions

137 Chapter 5: CASE EXAMPLES proved to be invaluable when applying the model to evaluate project alternatives. An additional data-related conclusion was that it is much better to use actual data rather than synthesized data, to the greatest extent possible. Actual data can include observed signal timing information when building networks, observed traffic flow properties when calibrating traffic flow parameters, and traffic counts and speeds when validating network model results. This effort also proved that it is not necessary to use matrix-estimation techniques to create input demand; this finding is significant because the use of matrix estimation in the context of future or alternative scenarios is problematic. A limitation of the current integration scheme that may be addressed in future model development phases is the lack of temporal information when extracting subarea demand. This effort also revealed to SFCTA staff a number of traffic microsimulation model sensitivities. For example, the traffic simulation model proved to be very sensitive to small changes in input assumptions. Staff described how a one-foot increase in the effective vehicle length caused systemwide network performance issues. Similarly, a bottleneck at a single intersection could also cause the entire network simulation to crash. Regarding using the model to evaluate alternative scenarios, staff discovered the following two points: The dynamic network model generally predicts more traf- fic diversion than static assignment techniques as a result of the sensitivity to actual capacity constraints, and stochasticity can be an issue when comparing scenarios, necessitating higher levels of convergence in order to draw meaningful conclusions. Finally, SFCTA staff advocated that sensitivity testing is an essential part of the model calibration and validation process (Parsons Brinckerhoff and San Francisco County Transportation Authority 2012). Next steps for model development include development of a disaggregate dynamic transit assignment model and better representation of parking behavior within San Francisco. Other key development tasks include the development of a full 24-hour simulation and the associated development of dynamic network-model-based skims for the entire day (Parsons Brinckerhoff and San Francisco County Transportation Authority 2012). MARICOPA ASSOCIATION OF GOVERNMENTS INNER LOOP TRAFFIC MODEL Objective The Maricopa Association of Governments developed the Inner Loop Traffic Model to support the Central Phoenix Transportation Framework Study. The purpose of this study is to identify the transportation strategies and investment needs for the central portion of the Phoenix region. The Phoenix core freeway system is still relatively new, but there are a significant number of chokepoints. Rather than only consider capac- ity expansion investments, the region wanted to have a tool that provided sensitivity to operational strategies in order to be able to more fully understand the interactions between the region’s highway system and regional arterials, and to strategically iden- tify how arterials can accommodate projected travel demand. In addition, significant

138 Part 2: ISSUES IN ADOPTING INTEGRATED DYNAMIC MODELS SYSTEMS investments in transit are being made by the region, and there is a tremendous focus around planning and developing high-capacity transit corridors. In order to have sen- sitivity to these strategies, it was determined that a more detailed network model than found in the trip-based demand model would be required. The development of the Inner Loop Traffic Model is considered the first part of a multiphase effort to develop regional simulation capabilities. This initial effort demonstrated the proof of the con- cept that it is feasible to develop and calibrate a regional-scale traffic simulation model (R. Hazlett, personal communication, Oct. 3, 2013). The information presented in this example was gathered from interviews with project team members. Model System Design and Components The focus of the Inner Loop Traffic Model development effort was to establish a large-scale network simulation model rather than to fully integrate the regional travel demand and network simulation components. The traffic model is only loosely linked with traditional trip-based travel demand model, and there is not yet an integrated demand–supply model runstream. Travel demand from the region’s trip-based model was used to seed the traffic simulation calibration effort, but the trip-based model demand was refined by applying a dynamic O-D matrix-estimation process that uses 15-minute counts. As a result, the present version of the model is more oriented toward shorter-term operations and engineering analyses rather than long-term future demand analyses. The project team is developing a process for creating detailed future-year simulation demand by pivoting off of the future demand model outputs. At present, network performance indicators, such as travel time and cost skims, are not being fed back to the trip-based demand model. The spatial extent of the simulation model is approximately 530 square miles. The model maintains a consistent geographic resolution with the trip-based model, including approximately 800 internal zones and 90 external or interface locations, but there are significantly more network loading locations than in the trip-based model. The original model design called for the simulation and calibration of two 3-hour peak periods, but ultimately the model included a broader temporal extent. This inclusion of the broader temporal extent was necessary to calibrate the 3-hour period that pre- cedes each of the peak time periods in order to ensure reasonable peak period network performance. This approach was especially necessary for the p.m. peak period. In the a.m. peak approximately 900,000 trips are simulated, while in the p.m. peak approxi- mately 1.2 million trips are simulated (D. Morgan, personal communication, Oct. 3, 2013). The project team was able to implement a microscopic model for the entire mod- eled area that incorporates the accurate representation of signals, meters, and bus routes and schedules. Use of a microscopic scale model provides better sensitivity to operational phenomena like traffic across lanes, weaving, merging, signals, and bottle- necks. The dynamic user equilibrium seeking solution method implemented in the model does not rely on the traditional method of successive averages as many dynamic network models do. Rather, at every iteration, every driver’s paths are informed by the latest network performance information. This approach is similar to the methodology

139 Chapter 5: CASE EXAMPLES used in the SHRP 2 C10A project. The model uses a temporal resolution of 15 min- utes for network pathbuilding, and a temporal resolution as fine as 0.1 seconds for the simulation step size (D. Morgan, personal communication, Oct. 3, 2013). The model development work was performed almost exclusively by consultants, although MAG staff received some training, and MAG has dedicated a staff person to ongoing model management. Agency staff are also now performing testing of the model to ensure that the tool is incorporating realistic assumptions and is producing reasonable results. The consultant selected to implement the model is a developer of one of the major traffic simulation software packages, and this familiarity with the software was one of the primary factors in selecting this consultant. Lessons Learned The most important lesson learned from this effort is that it is possible to build a regional-scale microscopic traffic simulation. Microscopic models can provide better sensitivity to operational phenomena such as traffic across lanes, weaving, merging, signals, and bottlenecks than mesoscopic models. Microscopic models have longer run times, but given operational considerations of interest to MAG, this trade-off was acceptable. Learning how to harness current hardware and software in order to achieve better run times was a key learning component of the project. An obvious lesson learned—but one that still bears repeating—is that regional simulation models require lots of good data and that it is preferable to use observed data rather than synthesized data. However, there are limitations on the availability of actual data; it is unavoidable that some assumptions and synthesis of data are necessary in establishing the model. In addition, calibration of the model system is challenging. Stochasticity, or random variation, was a particular focus of the team in develop- ment of the model. Stochasticity is intrinsic to the simulation model, as well as intrinsic to the real world, and this effort revealed there is significant investigation to be done to understand how the models can be run and how the results can be applied, interpreted, and communicated to member agencies and to the public. Two primary next steps are envisioned for this model. First, the regional net- work simulation is to be expanded to cover the entire region, rather than just the core 500 square miles. This work is currently ongoing. Second, MAG has also been developing a regional activity-based model system. Although the initial version of this activity-based model system incorporates a traditional static network assignment component, it is anticipated that at some point the activity-base demand model and regional traffic microsimulation model will be linked (R. Hazlett, personal communi- cation, Oct. 3, 2013). These case examples confirm the potential for integrated dynamic models to give more comprehensive and more detailed information to decision makers and to pro- vide sensitivity to a wider set of policy and investment alternatives. The C10A and C10B projects demonstrated these enhanced capabilities through a set of diverse policy sensitivity tests, while the SFCTA’s DTA Anyway project illustrated how an activity- based and dynamic network model model system could be used to inform real project choices. The SFCTA as well as the MAG efforts are also indicators of future directions

140 Part 2: ISSUES IN ADOPTING INTEGRATED DYNAMIC MODELS SYSTEMS in travel demand forecasting practice, as both agencies intend to move toward more fully integrated dynamic model systems with future model development efforts. How- ever, these case examples also illustrate issues (e.g., integration strategies and com- putational resource requirements) with developing an integrated model. In addition, implementing an integrated dynamic network model necessitates addressing the issues independently associated with implementing an activity-based model and a regional- scale dynamic network model. The following sections in Chapter 6 consider some of the critical issues faced when implementing integrated dynamic model systems.

Welcome to OpenBook!

You're looking at OpenBook, NAP.edu's online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

Do you want to take a quick tour of the OpenBook's features?

Show this book's table of contents , where you can jump to any chapter by name.

...or use these buttons to go back to the previous chapter or skip to the next one.

Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

To search the entire text of this book, type in your search term here and press Enter .

Share a link to this book page on your preferred social network or via email.

View our suggested citation for this chapter.

Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

Get Email Updates

Do you enjoy reading reports from the Academies online for free ? Sign up for email notifications and we'll let you know about new publications in your areas of interest when they're released.

Connect Socal

Trip-based model.

Existing Travel Demand Model

Trip-Based Travel Model: 4k

The trip-based model is a system of mathematical and statistical processes that estimates daily travel patterns within the Puget Sound region.

How Does the Trip-Based Travel Model Work?

For every household in the region, the model estimates how many trips are made each day, where they go, what time of day they travel, which modes they use, and which routes they follow. The relationships that are estimated for the 2014 base year are combined with future population, employment, and transportation infrastructure growth assumptions to produce future travel forecasts. The future travel forecasts are then analyzed to inform regional transportation studies and plans. The model uses Inro’s Emme software and is built entirely in the Emme macro language.

Reference files

• 4k Model Source Code
• Instructions for Version 4.0.3
• 2010 Model Inputs
• 2040 Model Inputs
• Version History—Version 4.0.3 Changes (Released: April 2015)
• Version 4.0.3 Documentation
• Travel Demand Forecasting at PSRC
• Value of Time Technical Memo

Fundamentals of Transportation/Agent-based Modeling

Agent-based modeling

Transportation engineers and planners rely on transportation forecasting models to address a wide range of increasingly complicated issues, from congestion and air quality, to social equity concerns. Two major strands of travel demand models have emerged over the past several decades, trip-based and activity-based approaches.

The traditional four-step travel demand model, often referred to as the trip-based approach, takes individual trips as the elementary subjects and considers aggregate travel choices in four steps: trip generation , trip distribution , modal split , and route assignment . This sequential travel demand modeling paradigm, which originated in the 1950s when limited data, computational power, and algorithms were available, ignores the diversity across individuals and does not have solid foundation in travel behavior theory. Discrete choice analysis , describes travel demand as a multi-dimensional hierarchical choice process, including residential and business location choice, trip origin, trip destination, travel modes and etc. Although discrete choice models could improve travel demand prediction by classifying travelers according to certain attributes such as age, gender and household incomes, it still in the end focuses on aggregate travel behaviors and ignores individual decision-making processes. Another flaw of four-step model lies in the fact that this sequential modeling process ignores the interaction between steps and can not predict certain phenomena such as induced travel or demand, which can be thought of as a feedback from traffic assignment to trip generation, distribution, and mode split. Although introducing feedback and iteratively applying four-step approach could mitigate this problem, researchers believe that a coherent framework should be introduced to address four steps simultaneously.

To overcome these inadequacies of conventional four-step modeling, activity-based models have been applied in travel demand analysis since the 1970s. Activity-based models predict activities and related travel choices by considering time and space constraints as well as individual characteristics. Individuals will follow a sequence of activities and make corresponding trips connecting those activities to maximize their utilities. Macroscopic travel patterns are predicted through aggregation of individual travel choice.

Although activity-based models have the potential to bridge the gap between individual decision-making processes and macroscopic travel demand, these models require solving many optimization problems simultaneously, which is computationally difficult and behaviorally unrealistic. Therefore, some models employ external aggregate methods such as User Equilibrium (Deterministic (DUE) or Stochastic (SUE)) to address route choice, which compromises their claims as microscopic decision-making models.

The agent-based travel demand models have emerged as a new generation of transportation forecasting tools and provide an alternative to addressing the topic of travel demand modeling. This modeling approach is flexible and capable to model individual decision making processes. There have been many applications of agent-based modeling in transportation (Transportation Research Part C (2002) dedicated a special issue to this topic). This modeling strategy, however, has not yet been widely adopted in travel demand modeling practice.

To build a pedagogically appropriate model, this chapter introduces an Agent-based Demand and Assignment Model (ADAM), extending Zhang and Levinson (2004), which addresses the destination choice and route choice problems with consideration of congestion. Students have the opportunity to work with the ADAM model for several exercises.

• 1 Introduction to agent-based models for transport
• 2.2.1 Destination selection rules: Network Origin-Destination Exploration
• 2.2.2 Path learning rule: Agent-based route choice
• 2.3 Iterations
• 3 Simulation
• 4 Further reading
• 5 References

Introduction to agent-based models for transport [ edit | edit source ]

While agent-based models are not commonly used in travel demand forecasting as such, many activity-based models are agent-based models of a sort, at least in part, though the behaviors of the agents are typically very complex. Historically, agent-based models come from different fields such as genetics, artificial intelligence, cognitive science, social science. The advantage of using them in transportation begins first with the intuition they provide. It makes more sense to people to think of individual travelers behaving rather than flows. This is in part because it is also more realistic, in that it can be formulated to capture the process by which travelers make decisions, and because it is tracking individuals, can be internally consistent (so that a given traveler has a particular set of constraints (like income, obligations, and time available)

There are several elements in an agent-based model:

• Agents are like people who have characteristics, goals and behavioral rules. The actions of agents depend on the environment they inhabit.
• The environment provides a space where agents live. The environment is shaped by the actions of agents.
• Interaction rules describe how agents and the environment interact

An agent-based model evolves by itself once those micro-level elements are specified. Macro-level properties emerge from this evolutionary process.

An exploratory agent-based model is presented below. The advantage of this model is its simplicity. Clearly, it will lose some predictive detail, but hopefully gives you a flavor of the kinds of modeling approaches and things that can be modeled with agent-based models in the realm of travel demand.

Agent-based Demand and Assignment Model (ADAM) [ edit | edit source ]

The agent-based modeling approach assumes that aggregate urban travel demand patterns emerge from multi-dimensional choice process of individuals. All agents have individual characteristics, goals, and rules of travel behaviors. Agents exchange information with the environment on their travel experiences and adjust their travel choices according to available information. In ADAM travelers are active agents and nodes are fixed point agents, while links comprise the environment.

ADAM can be thought of as modeling the AM commute. As shown in Figure 1, ADAM examines the status of each traveler after updating turning matrices at nodes. If a traveler has not found a satisfactory job (status = 1), that traveler will continue the random process of job searching following the rules presented later in this paper. The process will repeat until either all travelers have found jobs (chosen a destination) or some maximum number of iterations are reached. The key components of the agent-based model are introduced in turn below.

Agents [ edit | edit source ]

Travelers aim to find a job on the network and a route leading from their origin to this destination with the lowest cost. In the searching process, each traveler visits a node and decides to either accept or reject a job available at that node according to rules discussed later in this paper. If they reject a job at that node, they proceed to another node. Travelers learn current link travel times in the neighborhood of the node when they visit a node through this link and they will only proceed through one link at each step. By accumulating link travel time information during the trip, travelers could derive travel cost between any two nodes they visited.

Nodes are geographic locations where links intersect in the real world. In this model, they also represent the abstract centroids of traffic zone where travelers originate from and are destined to. Furthermore, nodes are carriers of pooled, collective knowledge, including both shortest path information and attractiveness of adjacent nodes. Travelers would exchange knowledge with nodes once they arrived at a new node. The knowledge and exchanging behavior is an abstraction of information spread in a community and communications among travelers in the real world. Links represents roads in the real world and have attributes such as length, free flow travel time, and capacity. Links also provide information about traffic flow and travel time to travelers passing by, which abstracts travelers’ observation of traffic condition in the real world. Links impose geographic constraints to travelers since they are only able to visit adjacent nodes directly connected by a link with the node they are currently visiting.

Rules [ edit | edit source ]

Rules are the most important attributes of an agent-based model, which drives the evolution of the model given initial condition. There are two fundamental rules in ADAM: turning rules for finding a destination and information exchange rules for improving paths.

Destination selection rules: Network Origin-Destination Exploration [ edit | edit source ]

The first element of ADAM is for each traveler, the discovery of a destination. The model which does this Network Origin-Destination Exploration (NODE) is described below.

Different definitions of turning probability reflect assumptions of different underlying decision-making processes of travelers regarding where to work and may lead to very different travel demand patterns on the network. Zhang and Levinson (2004) assumed that this probability is proportional to jobs available at each node and ignored the ease of reaching them (travel cost). Another disadvantage of this assumption is that if a node does not have available jobs, travelers will never search in this direction even though more jobs may be available via this node.

Path learning rule: Agent-based route choice [ edit | edit source ]

The other important rule in ADAM is the path learning rule. Travelers will learn travel cost of links on their travel route, while nodes keep information about the shortest path from itself to all other nodes which have been visited by travelers to that node. Once a traveler arrives at a new node, that traveler compares their knowledge about travel cost from the current node to each node on the traveler’s travel route. Both of them will keep the shorter "shortest path" after knowledge exchange. Although nodes originally have very limited knowledge about the routes in the remaining network, information spreads rapidly on the network. With the congested link travel time, which can be simply defined as any available travel time-flow relationship, each traveler’s choice will change the link travel time on the network and thus affect destination and route choice of other travelers. Travelers’ route adjustments will trigger more significant change on the network thus other travelers’ behavior. This mechanism reflects the complexity of the real world.

ADAM's Agent-based route choice component (ARC) simulates individual route choices and determines the flow pattern on the network subject to a given OD distribution.

The initial route choice can be either given or generated by a random-walk route searching process at iteration 0. In the random walk scenario, travelers set off from their origins and travel in a randomly chosen direction, updating directions after arriving at each node. However, directed cycles and U-turns are prevented. Once travelers arrive at the destination, their travel routes become the initial travel route and will be updated in subsequent iterations. The randomness of searching direction and the large number of travelers will ensure the diversity of initial route choices, which comprises the knowledge based on subsequent iterations.

On subsequent iterations, each traveler follows a fixed route chosen at the end of the previous iteration. Once arriving at a destination centroid, travelers will enrich the information set with their individual knowledge while benefiting from the pooled knowledge at the same time by exchanging both shortest path and toll information with centroids. Those travelers will also bring that updated information back to their origin and repeat the exchange process. The information exchange mechanism is illustrated by Figure 1.

As illustrated in Figure 1, suppose that the traveler originating at node 1 is traveling to node 5, initially via node 4. His initial shortest path knowledge is 1-3-4-5. Suppose the shortest path information stored at node 5 is 4-5, 3-5, 2-3-5 and 1-2-3-5, respectively from nodes 4, 3, 2 and 1. The comparison starts from the node closest to the current node along the path chain in traveler's memory and repeats for each node on this chain until reaching the origin. After comparing the path from node 3 to 5, the traveler's path information is updated to 1-3-5 since the shortest path for this path segment proposed by the node is shorter than that held by the traveler. Notice that this improvement has also changed the shortest path from node 1 to 5 in the traveler's memory. Consequently, the node will adopt the path from node 1 proposed by the traveler since 1-3-5 is better than 1-2-3-5. The updated path from node 1 to 5 then becomes part of the traveler's shortest path information. This information exchange mechanism will naturally mutate the path chain and generate the most efficient route, sometimes better than all known existing routes. Since nodes store K alternative paths, nodes will insert the path proposed by the visitor in their information pool as long as this path is better than the longest path stored. This information will also be shared with those travelers visiting node 5 at subsequent steps.

After stopping at the destination node, travelers compare their travel route determined at the end of previous iterations and shortest path learned during the currently iteration. The path length is evaluated in dollar value by each traveler, considering their individual value of time and the toll charged by each link segment. Since travelers have different values of time, the cost of K alternatives should be reevaluated and sorted for each traveler. If the path suggested by the destination node is better than their current route, the travelers have a probability to switch to the better route that iteration. In general,

To apply this model, we choose a specific form:

ARC simulates the day-to-day route choice behavior of travelers and this probability curve must account for two factors:

• the probability a traveler perceives this better path once its information is available and
• the probability a traveler takes this path once it is learned. It should be noted that information spreading takes time and not everyone learns immediately.

Travelers with more effective social networks are more likely to be exposed to such information and thus have a higher probability of learning the better path. Once a new road opens, it takes weeks or even months before the flow reaches a stable level. Even when people learn a better alternative, route change involves a certain switching cost preventing travelers from changing routes immediately. Or travelers may just resist changing because of inertia. Considering these factors, this curve should increase as benefits increase and reach some upper limit predicted by the willingness to learn. Estimation of this curve through survey or other psychological studies will enhance the empirical foundation of the model.

Iterations [ edit | edit source ]

Traditional travel demand models disentangled this complexity by formulating an optimization problem, using either Deterministic or Stochastic User Equilibrium. However, algorithms employed to solve such optimization problems are computationally cumbersome and behaviorally unrealistic. Instead, ADAM introduces a heuristic learning process to address this challenge. Under this framework, travelers will reenter the network and choose their destination and route again according to the link travel time resulting from their previous choices. Updated shortest path information will be learned and spread by travelers. This process mimics people’s job change and route change behavior. Given the initial condition, ADAM evolves with previously defined rules and a pattern may be achieved according to certain convergence rules, from which macroscopic information such as trip distribution and traffic assignment can be extracted by summing up individual choices.

Simulation [ edit | edit source ]

• Agent-based Demand and Assignment Model interactive model

Note this software takes the number of trips, the share of those trips by automobile, and the number of trips in the peak hour as given exogenously by the user. More complex agent-based models could consider those directly. The arcs (links) in the model estimate travel time using a link performance function, described in Route Choice

Further reading [ edit | edit source ]

• A Primer for Agent-Based Simulation and Modeling in Transportation Applications published by Federal Highway Administration.
• Activity-Based Travel Demand Models: A Primer published by Strategic Highways Research Program.
• A Transportation Modeling Primer by Edward A. Beimborn
• Zhu, Shanjiang and Levinson, David (2018) Agent-Based Route Choice with Learning and Exchange of Information. Urban Science 2(3), 58 .
• Di, Xuan, Henry Liu, and David Levinson. (2015) Multi-agent Route Choice Game for Transportation Engineering. Transportation Research Record 2480 55-63 .
• Tilahun, Nebiyou and David Levinson (2013) An Agent-Based Model of Worker and Job Matching. Journal of Transport and Land Use 6(1) 73-88 .
• Huang, Arthur and David Levinson (2011) Why retailers cluster: An agent model of location choice on supply chains. Environment and Planning b 38(1) 82 – 94 .
• Zhu, Shanjiang, Feng Xie and David Levinson (2011) Enhancing Transportation Education through On-line Simulation using an Agent-Based Demand and Assignment Model. ASCE Journal of Professional Issues in Engineering Education and Practice 137(1) 38-45 .
• Zhang, Lei, David Levinson, and Shanjiang Zhu (2008) Agent-Based Model of Price Competition and Product Differentiation on Congested Networks. Journal of Transport Economics and Policy Sept. 2008 42(3) 435-461 .
• Zou, Xi and David Levinson (2006) A Multi-Agent Congestion and Pricing Model. Transportmetrica 2(3) 237-249 .
• Zhang, Lei and David Levinson. (2004a) An Agent-Based Approach to Travel Demand Modeling: An Exploratory Analysis. Transportation Research Record: Journal of the Transportation Research Board 1898 28-38 .
• Zou, Xi and David Levinson (2003) Vehicle Based Intersection Management with Intelligent Agents. ITS America Annual Meeting Proceedings .

References [ edit | edit source ]

• Bar-Gera, Hillel, 2001, Transportation Network Test Problems, Ben-Gurion University of Negev, http://www.bgu.ac.il/~bargera/tntp/
• Ben-Akiva, M. and Lerman S.R., 1985, Discrete choice analysis: theory and application to travel demand. The MIT Press, Cambridge, Massachusetts
• Boyce, D., 2002, Is the sequential travel forecasting paradigm counterproductive. ASCE Journal of Urban Planning and Development 128(4): 169-183
• Handy S., 1993, Regional versus local accessibility: Implications for non-work travel. Transportation Research Record 1400: 58–66.
• Handy, S., et al. 2002. Education of Transportation Planning Professionals. Transportation Research Record, no. 1812, pp. 151–160.
• Kitamura, R., Pas, E.I., Lula, C.V., Lawton, K. and Benson, P.E., 1996, The Sequenced Activity Mobility Simulator (SAMS): An Integrated Approach to Modeling Transportation, Land Use and Air Quality. Transportation 23: 267-291
• McFadden, D., 1974, The measurement of urban travel demand, Berkeley : Institute of Urban & Regional Development, University of California.
• Parthasarathi, P., Levinson, D., and Karamalaputi, R., 2003. Induced Demand: A Microscopic Perspective. Urban Studies, Volume 40, Number 7, pp 1335–1353.
• Pas, E.I., 1985, State of the art and research opportunities in travel demand: another perspective. Transportation Research Part A(21): 431-438
• Recker, W.W., McNally, M.G., and Root, G.S., 1986, A model of complex travel behavior:part I-theoretical development. Transportation Research 20A: 307-318
• Kitamura, R., 1988, An evaluation of activity-based travel analysis. Transportation 15: 9-34
• Timmermans, H.J.P., Arentze, T.A. and Joh, C.-H., 2002, Analyzing space-time behavior:new approaches to old problems, Progress in Human Geography, 26, 175-190.
• Zhang, Lei and David Levinson. 2004. An Agent-Based Approach to Travel Demand Modeling: An Exploratory Analysis. Transportation Research Record: Journal of the Transportation Research Board #1898 pp. 28–38

• Book:Fundamentals of Transportation

Navigation menu

Destination choice models

Introduction

Advantages and Limitations of Destination Choice Models

Destination Choice Models in Practice

Destination Choice: Theoretical Foundations

Destination Choice: Mathematical Formulation

Destination Choice Set Formation

Joint/Conditional Destination Choices

Factors Affecting Destination Choice

Destination Choice: Data Sources

Destination Choice: Parameter Estimation

Destination Choice: Calibration and Validation

Destination Choice: Implementation and Application

Page categories

Spatial Interaction Models

Topic Circles

More pages in this category:

# introduction.

(opens new window) , Tennessee; Charlottesville, Virginia; Charleston, South Carolina; and Jacksonville, Florida.

Destination choice models are a type of trip distribution or spatial interaction model which are formulated as discrete choice models , typically logit models. They can be thought of as a generalization of the traditional and widely used gravity model. In practice, this flexible and extensible formulation allows destination choice models to provide a better behavioral basis for trip distribution than the traditional gravity models, by allowing for a wider range of explanatory variables. Although technically gravity models can be considered a subset or special case of destination choice models, the term “destination choice models” typically is used to identify models that incorporate additional variables beyond size/attractions, impedance/friction factors and constants or k-factors. (see, for example, [3] [4] [5] [6] [7] [8] ).

# Advantages and Limitations of Destination Choice Models

Although the gravity analogy may still be appropriate for trip distribution in mono-centric urban regions where accessibility to transit plays little to no role in choice of destination, this is no longer the case in many urban areas, where there may be more than one dominant attraction region, multiple and important suburban-to-suburban trip flows, and where there is interest in understanding the contribution of transit towards achieving more sustainable urban development patterns. The gravity model often times exhibits incorrect demand elasticities; in particular, the model may respond illogically to changes in levels of service where improved accessibility to a given destination may cause a disproportionate increase in total trips, and/or an increase in trips using the mode(s) whose accessibility did not change. Both of these results are undesirable and may lead to erroneous assessments of the impact of transit or highway improvements. Destination choice models can overcome these gravity model limitations. With appropriate specifications of utility, consistency between changes in levels of service and changes in demand can be assured when using well validated destination choice models. In addition, because the functional form of the destination choice utility is very flexible, accounting for singularities in the trip distribution pattern can be accomplished in intuitive ways. For example, to inform a gravity model of the presence of a natural barrier, such as a river, a K-factor is often used. With a destination choice model, a term can be added to the utility equation, statistically estimated from observed data, and interpreted in terms of equivalent minutes of travel time; a much more data-based and intuitive measure of the impact the river would have on a person's travel choice.

While a key advantage offered by destination choice models when compared to the more traditional gravity model is their ability to consider additional factors, at the same time it is also important to recognize destination choice models in practice today still struggle to explain the spatial distribution of personal travel. This is due in large measure to the the importance of unobserved attributes such as the price and quality of goods and services provided at destinations. In many cases, a destination choice model may be able to double the goodness-of-fit, or explain twice as much of the observed travel patterns than a gravity model, but in the end still explain less than half of the variation in the observed patterns.

Both the advantages and limitations of destination choice models can be understood in terms of the factors that affect travelers' destination choices based on those the models can incorporate or reflect and those they cannot. The table below describes some of these advantages and limitations.

# Destination Choice Models in Practice

Destination choice models can be used in aggregate trip-based models as an alternative to gravity models or other spatial interaction models. Destination choice models are standard and ubiquitous in tour-based and activity-based models.

(opens new window) , 9% of MPOs & DOTs were using a tour-based or activity-based model and an additional 17% were in the process of developing them. Destination choice models are therefore likely currently in use in approximately 15% of travel models and likely to be used in roughly a third of models in the relatively near term future.

(opens new window) for a statewide example.)

For tips on things to check when developing or using a Destination Choice Model, check out this page .

# Destination Choice: Theoretical Foundations

Destination choice (and trip distribution) models can be derived from a variety of theoretical starting points, the most common of which are entropy maximization and random utility theory. Both approaches generate mathematically similar functional forms that can generally be classed as spatial interaction models. These theoretical foundations define the basic assumptions of destination choice models, their functional forms and parameter estimation requirements.

# Destination Choice: Mathematical Formulation

The most common mathematical formulation of destination choice is the multinomial logit (MNL) model. Gravity models, which are commonly used in aggregate, trip-based models, can be shown to be a special case of a MNL destination choice model. [10] Another type of early trip distribution model is the intervening opportunities model, [11] but this model has fallen into disuse in North America. On the other hand, data-driven approaches are emerging, facilitated by the availability of passive origin-destination big data.

# Destination Choice Set Formation

Choice set formation is a critical step in the specification, estimation, and application of all discrete choice models, including destination choice. As noted by Thill (1992), the misspecification of choice sets can have adverse effects on parameter estimates and resultant computations of predicted choice probabilities. The accurate definition of the destination choice set has been an issue of much interest to the profession and a variety of approaches have been developed and adopted in research and practice. The problem is of acute significance in the context of destination choice modeling because the number of elemental alternatives can be very large. With many travel demand model systems comprising thousands of zones, destination choice sets can prove to be extremely large. On the one hand, methodological and computational advances now allow the use of the universe of locations as the destination choice set. On the other hand, the use of universal set of destinations as the choice set may compromise the behavioral representativeness of destination choice models. The analyst needs to consider the pros and cons of alternative approaches carefully when defining destination choice sets.

# Joint/Conditional Destination Choices

In traditional four-step travel models, all destination choices are assumed to be independent. This is reflected in the fact that gravity or destination choice models in this context are run in parallel and independently of each other. Both activity-based, hybrid, and advanced trip-based frameworks have developed different approaches to relaxing this assumption of independence. Each of these approaches is presented here .

# Factors Affecting Destination Choice

Typically, zone-based destination choice models will incorporate a utility function that includes two categories of explanatory factors: qualitative factors (how good are the choices in a given destination zone), and quantitative factors (how many individual choices are in a zone). The usage of qualitative explanatory factors is common in virtually all choice models. For destination choice models, these commonly include impedance , accessibility , psychological boundaries , and other destination qualities , as well as traveler attributes . The quantitative factors, typically labeled as size terms or attractions , are an unusual feature of destination choice models, which arise because the "alternatives" represented in the model, often TAZs , are not actually the choices, but they represent a pool of choices. The actual choice is instead one particular activity point (job, store, theatre seat, etc.) within the zone. Due to this distinction, factors that represent the quantity (instead of quality) of choices in a zone need to be treated differently in the mathematical formulation, as documented here . A more extensive description of the various factors in destination choice models is found here .

# Destination Choice: Data Sources

The flexibility of destination choice models comes at a cost. While it is possible to represent the selection of trip destinations more rigorously, destination choice models tend to require more data and data with higher fidelity than traditional [. There are two types of data that are relevant for destination choice models. Observed choice data describe origin-destination flows that have been observed in a survey, by counting or by passive data collection. Explanatory data, on the other hand, refer to input data that describe either destinations or characteristics of the decision maker who chooses the destination. Further discussion on data sources is available here .

# Destination Choice: Parameter Estimation

Once observed choices and explanatory variables from data are related by formulating a utility function, the challenge becomes estimating the parameters that quantify these relationships or how explanatory variables contribute to destination choice probabilities. Rather than a one time effort, this is commonly an iterative process in which alternative specifications of the utility function are tested. The parameter estimation process is based in statistical / econometric theory and generally relies on maximum likelihood estimation (MLE) techniques. Specialized software or custom programming is generally required. Algorithmic approaches to MLE for destination choice models generally fall into two general families: gradient-based and metaheuristics. For more information see here .

# Destination Choice: Calibration and Validation

In practice, destination choice models can rarely be applied for forecasting exactly as they are estimated. Calibration adjustments are commonly required for several reasons. Sometimes application of the model to application data sets produce results that differ in some important ways from the results when the model is applied to the estimation data sets. In some cases such differences can be caused or exacerbated by inconsistencies between the model estimation and application (such as different sources for explanatory variables like income or travel time or the omission of constraints in estimation). Careful and thoughtful adjustments in keeping with good professional judgment can be required in order ensure the applied model demonstrates both reasonable ability to replicate observed travel patterns (from both estimation data and in some cases, other independent data sources for validation) and reasonable response properties or elasticities to key variables. A variety of measures can be used to evaluate the validity of destination choice models. Comparisons to trip length frequency distributions remain the most common approaches although it has been demonstrated that models can easily be over-calibrated to reproduce trip length frequency distributions at the expense of their ability to accurately reproduce actual spatial interaction patterns. [12] Various strategies can be and are commonly employed such as adjustments to distance or impedance parameters , the assertion of size or mode choice logsum parameters, and the use of constants. For more information on calibration and validation see here .

# Destination Choice: Implementation and Application

Destination choice models can be implemented in various ways in different travel modeling frameworks. They can be applied disaggregately in activity-based models using Monte Carlo simulation or aggregately in trip-based models using matrices. In both of these contexts, there are important issues related to how destination choice models are integrated with the larger model system. Key issues include how various destination choices are related to each other, how choices of destination and mode are related to each other, and how the larger model system acheives an equilibrium between travel demand and supply, commonly through iterative feedback loops. For more in-depth discussion, click here .

# References

• Content Charrette: Destination Choice Models

Zhao, Y. and K. Kockelman (2002) 'The Propagation of Uncertainty through Travel Demand Models', Annals of Regional Science 36 (1), pp.145-163 ↩︎

Bernardin, V. L., F. Koppelman, and D. Boyce. Enhanced Destination Choice Models Incorporating Agglomeration Related to Trip Chaining While Controlling for Spatial Competition. In Transportation Research Record: Journal of the Transportation Research Board of the National Academies, Washington, D.C., 2009, pp. 143-151. ↩︎

Chow, L.-F.,, F. Zhao, M.-T. Li, and S.-C. Li. Development and Evaluation of Aggregate Destination Choice Models for Trip Distribution in Florida. In Transportation Research Record: Journal of the Transportation Research Board of the National Academies, Washington, D.C., 2005, pp. 18-27 ↩︎

Jonnalagadda, N., J. Freedman, W. A. Davidson, and J. D. Hunt. Development of Microsimulation Activity-Based Model for San Francisco: Destination and Mode Choice Models. In Transportation Research Record: Journal of the Transportation Research Board of the National Academies, Washington, D.C., 2001, pp. 25-35. ↩︎

Bhat, C., A. Govindarajan, and V. Pulugata. Disaggregate Attraction-End Choice Modeling: Formulation and Empirical Analysis. In Transportation Research Record: Journal of the Transportation Research Board of the National Academies, Washington, D.C., 1998, pp. 0-68 ↩︎

Borgers, A., and H. Timmermans. Choice Model Specification, Substitution and Spatial Structure Effects: A Simulation Experiment. Regional Science and Urban Economics, Vol. 17, 1987, pp. 29-47 ↩︎

Fotheringham, A. S. Some Theoretical Aspects of Destination Choice and Their Relevance to Production-Constrained Gravity Models. Environment and Planning, Vol. 15A, 1983, pp. 464-488 ↩︎

SR 288-Metropolitan Travel Forecasting Current Practice and Future Direction ↩︎

Daly, A. Estimating Choice Models Containing Attraction Variables. Transportation Research B, Vol. 16B, No. 1, 1982, pp. 5–15. ↩︎

Stouffer, S. Intervening Opportunities: A Theory Relating Mobility and Distance. American Sociological Review, Vol. 5, 1940, pp. 845–867. ↩︎

Ye, X., W. Cheng and X. Jia (2012) "Synthetic Environment to Evaluate Alternative Trip Distribution Models." Transportation Research Record: Journal of the Transportation Research Board, No. 2302: p. 111-120. ↩︎

← Trip distribution Mode choice →

This site uses cookies to learn which topics interest our readers.