National Academies Press: OpenBook
« Previous: Chapter 2 - Travel Demand Metrics
Page 14
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 14
Page 15
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 15
Page 16
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 16
Page 17
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 17
Page 18
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 18
Page 19
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 19
Page 20
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 20
Page 21
Suggested Citation:"Chapter 3 - TDFM Updates." National Academies of Sciences, Engineering, and Medicine. 2024. New Mobility Options in Travel Demand Forecasting and Modeling: A Guide. Washington, DC: The National Academies Press. doi: 10.17226/27827.
×
Page 21

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.

14 TDFM Updates The performance metrics identified in the preceding chapter will provide useful metrics that can help practitioners examine whether travel demand in the jurisdiction has been altered by NMO adoption. This chapter will provide direction to transportation agencies on how to implement TDFM updates for the selected components. Based on the performance metrics, one component or multiple components can be selected for updating. The discussion is organized to present the research philosophy employed to illustrate model update procedures for three use case examples. The use case examples cover trip-based and tour-/activity-based model systems. 3.1 TDFM Update Considerations Analysis of travel demand metrics provides agency personnel the set of components that need to be updated in response to the adoption of NMOs. At this stage, it is important to recognize the scale of the component update process. Depending on the specific component selected, the scope of work for a TDFM update could vary significantly. The main changes to be considered in the update process will be (a) data compilation and variable customization, (b) re-estimation of the same model functional form for new data, and (c) potential changes to the modeling structure due to NMO presence. When agency personnel are deciding whether to update a TDFM model, they should carefully consider the resources that would be needed. Consider a scenario where an examination of the vehicle ownership metric identified signifi- cant changes across the different alternatives. The changes in vehicle ownership do not require the consideration of a new mathematical model structure. The same dependent variable form, such as a multinomial logit model of the previous TDFM component, can be retained. However, in this scenario, the update process will require the compilation of data for the post-NMO scenario. Then, the data compiled can be used to re-estimate the vehicle ownership model with a host of independent variables related to the study region and NMO adoption. The generation of these variables will be time consuming and will need to be considered in the scope of work. Consider changes to travel mode choice distributions in the study region. With the emergence of NMOs, it is likely that the share of NMOs is significant in the study region. This is an example of a change in the number of alternatives of the mode choice model. The update process will be resource intensive because the analyst will need to compile new mode choice datasets, establish a new math- ematical structure with a larger number of alternatives, and then re-estimate the model specification. 3.2 Framework for Update Process NMO-related TDFM update implementation is still in its infancy. Hence, there is a lack of data at the jurisdiction level to provide use case guidance for NMOs. The research team iden- tified a solution to address this challenge. The research team has used real-world datasets that are C H A P T E R   3

TDFM Updates 15   currently employed for base model development and adapted them to reflect potential NMO- related scenarios. Let us consider a TDFM where 2017 data were employed for model estimation and calibration by a jurisdiction. The 2017 data would not have reflected an adequate share of shared micromobility or TNCs. The research team systematically synthesized an updated future dataset that reflects the growing proportion of shared micromobility and TNCs. While the dataset is not real, it is embedded with synthesis logic that will still result in plausible records. For example, in mode choice data, the researchers modified a specific number of trips to reflect TNCs instead of car or carpool alternatives. The team documented the process employed for updating the mode choice in Section 3.5: Use Case Example 3: Trip Mode Choice Model. In this manner, a real-world–like dataset will be ready for TDFM update consideration. Using this synthetic data, the research team will illustrate how transportation practitioners can update the appropriate TDFM component. A sample illustration of the data is provided in Tables 3.1 and 3.2. The synthetic data generation allows the researchers to circumvent the issues related to the absence of real-world data. When real-world data are available, jurisdictions can simply apply the approaches described using their jurisdiction-specific dataset. Thus, the guide’s suggestions will remain relevant in the future. The base data, synthesized data, base model, and updated model will illustrate how the TDFM update process was carried out. The research team has provided the scripts employed for data processing and model calibration, providing jurisdic- tions with real-world, step-by-step procedures for TDFM updates (see the appendix). A consideration of model update procedures for all TDFM components is beyond the scope of this project. Hence, to reflect model heterogeneity and the complexity of the TDFM update process, the authors selected the following three components for use case examples: (a) house- hold vehicle ownership, (b) household trip rates, and (c) trip-/tour-level mode choice. For these three use case examples, the authors have documented the model update procedures. In the first step, the authors have built a model specification for observed real data in the absence of NMOs. As NMOs are introduced across jurisdictions, there could be a change in travel demand because Real-World Commute Mode Data (2015) Household No. Person No. Commute Mode Time (mins) Cost ($) … Number of Cars 1 1 Car 15 1.5 … 3 1 2 Car 6 0.5 … 3 2 1 Transit 25 0.5 … 1 2 2 Walk 15 0 … 1 … … 15 2 Car 12 1.0 … 3 Synthesized Commute Mode Data (2025) Household No. Person No. Commute Mode Time (mins) Cost ($) … Number of Cars 1 1 Car 15 1.5 … 3 1 2 TNC 6 0.5 … 3 2 1 Transit 25 0.5 … 1 2 2 Shared Micromobility 15 0 … 1 … … 15 2 Car 12 1.0 … 3 Note: The records updated are in bold. Table 3.1. Observed commute mode choice data for 2015. Table 3.2. Synthesized commute mode data for 2025.

16 New Mobility Options in Travel Demand Forecasting and Modeling: A Guide of the NMO addition, or there could be a change in travel demand without any correlation to NMOs. To reflect these two possibilities, the authors have built models for two simulated scenar- ios. The first simulated scenario represents a modification of the dependent variable (from real data) in response to NMO introduction. The second simulated scenario represents a modifica- tion of the dependent variable over time without any significant impact from NMO addition. The two newly simulated datasets are employed to update the original model specification developed with real data to illustrate the two possibilities of NMO introduction across jurisdictions. 3.3 Use Case Example 1: Household Vehicle Ownership For this use case example, the research team developed three different vehicle ownership models using data from the 2011 Atlanta Household Travel Survey (Household Travel Survey, 2011) and auto ownership model specifications from the Atlanta Regional Commission (ARC) travel demand model (ARC Model Specification, 2015). The research team estimated a multino- mial logit (MNL) model considering four alternatives in vehicle ownership level: (a) households (HHs) with 0 vehicles, (b) HHs with 1 vehicle, (c) HHs with 2 vehicles, and (d) HHs with 3 or more vehicles. Vehicle-ownership–dependent variables were employed for model development. The house- hold vehicle ownership variable in Model A (see Table 3.3) is based on the 2011 ARC data and explicitly recognizes that NMOs were not available in the jurisdiction at the time. The vehicle ownership variable in Model B was compiled by modifying the 2011 household variable with the assumption that access and availability of NMOs reduced vehicle ownership. The vehicle ownership variable in Model C was compiled by modifying the 2011 household variable with the assumption that NMOs did not have an impact on vehicle ownership. A comparison of the vehicle ownership levels for the three scenarios is presented in Table 3.3. From Table 3.3, it can be concluded that vehicle ownership levels are lower in the scenario where NMOs affect vehicle ownership (see the increased share of 0 cars and the reduced share of 3+ cars). The fourth column representing vehicle ownership rates unaffected by NMOs offers very distinct ownership levels relative to the base year. This column shows higher 0-car and 3+ car ownership levels. Table 3.4 presents the MNL model estimation results for the data pre- sented in Table 3.3. The first column presents the variables considered; the second, third, and fourth columns present the parameter estimates (with their t-statistics in parentheses) for the base model scenario, the scenario where NMOs have an impact, and the scenario where NMOs have no impact, all for vehicle ownership levels of 0, 1, and 2. For model development, the 3+ ownership level was considered as the base category. Based on this base category, in Table 3.4, a positive (negative) sign of an independent variable for an ownership alternative represents an increased (decreased) propensity to fall in the corresponding vehicle ownership alternative compared to the 3+ ownership level. The reader can see that the model estimation results across all three models are intuitive. The results also show that NMO variables do not affect models for base data and the Model C data. Vehicle Ownership Base Model A (%) Model B NMOs Had an Impact (%) Model C NMOs Did Not Have an Impact (%) 0 4.2 11.6 10.6 1 25.2 25.8 23.6 2 42.8 39.6 33.3 3 or more 27.8 23.1 32.5 Table 3.3. Share of vehicle ownership.

TDFM Updates 17   Variable Base Model A Model B NMOs Had an Impact Model C NMOs Did Not Have an Impact Vehicle Ownership Level 0 1 2 0 1 2 0 1 2 Constant –2.343 –1.317 0.338 –3.310 –1.195 0.293 –1.689 –0.869 –0.068–(11.704) –(17.395) (10.799) –(17.341) –(16.003) (8.943) –(17.466) –(18.142) –(2.144) Ratio of workers to drivers –2.161 0.357 — –0.889 0.179 — –0.400 — — –(12.112) (5.556) — –(7.212) (2.803) — –(4.902) — — Ratio of people aged 6–15 to drivers — –0.160 0.382 — — 0.412 — — 0.227— –(2.095) (6.490) — — (8.240) — — (4.610) Ratio of people aged 18–24 to drivers –1.704 –2.420 –2.067 — — — –2.001 –1.574 –1.561 –(5.263) –(12.981) –(12.902) — — — –(8.018) –(9.060) –(9.530) Ratio of people aged 80 or older to drivers — 0.484 — — 0.439 — — 0.322 — — (5.507) — — (5.082) — — (3.754) — HH income <=$10,000 5.254 4.076 1.257 5.547 3.467 0.952 3.913 2.764 1.374(15.528) (14.615) (4.279) (16.885) (12.517) (3.258) (16.797) (12.380) (5.922) $10,000 < HH income <= $30,000 3.379 2.886 0.677 3.645 2.491 0.437 2.398 1.850 0.628(15.223) (26.193) (6.524) (17.150) (22.384) (4.106) (19.759) (19.468) (6.770) $30,000 < HH income <= $60,000 1.069 1.784 0.407 1.919 1.538 0.191 1.131 1.099 0.298(4.299) (22.003) (6.204) (9.058) (18.323) (2.746) (10.294) (14.860) (4.649) $60,000 < HH income <= $100,000 –0.760 0.518 — — 0.388 — 0.381 0.168 —–(2.079) (7.164) — — (5.740) — (3.561) (2.471) — NMO availability indicator (urban HH indicator) — — — 3.817 1.637 1.247 — — — — — — (19.493) (10.461) (10.708) — — — NMO availability * presence of worker in HH — — — 0.905 — — — — — — — — (5.359) — — — — — NMO availability * household income >= $60,000 — — — 1.019 0.811 — — — — — — — (4.341) (5.566) — — — — Note: Dashes (—) indicate that these variables were not found to be significant at 90% confidence level (t-statistic 1.645) Table 3.4. Vehicle ownership model estimation results. However, it is important to note that the model specification varies significantly from base data to Model C data. These results indicate that while NMOs did not affect vehicle ownership, the model parameters did change for other variables from 2011 under the data generation assumptions for Model C. Finally, as intended, in Model B, it can be observed that NMO access and interactions of access with workers and income highlight a higher utility for lower vehicle ownership levels. Across these variables, the actual ownership levels affected by the variables are different. In the appendix, the research team has included the data and associated codes for model devel- opment in R format so that practitioners can test out the code and approaches to plan for specifica- tion testing. When jurisdiction-specific data are available, practitioners can replace the columns of simulated dependent variables and apply these model codes for jurisdiction-specific model development. The overall structure of updating for vehicle ownership is shown in Figure 3.1. Potential changes to the model componentsUpdates to considerUse Case Example 1 Household vehicle ownership levels Change in independent variables Update MNL model specification for traditional variables Include NMO variables Test NMO-specific variables to verify NMO impacts Figure 3.1. Updates for household vehicle ownership model.

18 New Mobility Options in Travel Demand Forecasting and Modeling: A Guide 3.4 Use Case Example 2: Household Trip Rate Model HH trip rate is an important variable of interest and is an important component of the trip- generation module. In an effort to identify the potential impact of NMOs on HH trip rate, the authors present a simulated scenario using 2011 Atlanta household survey data (Household Travel Survey, 2011) and ARC travel demand model specifications (ARC Model Specification, 2015). The research team developed three different models: (a) Model A, a base trip-generation– rate model considering t he absence of NMOs in the region using 2011 Atlanta household survey data; (b) Model B, an updated trip-generation–rate model with the presence of NMOs in the region where NMO availability affects trip generation; and (c) Model C, an updated trip- generation–rate model where NMO availability does not affect trip generation, but the model parameters are revised. For Model B, the trip rate data were generated by modifying the 2011 household trip- generation data to account for the impact of NMO availability on household trip rates. For Model C, household trip-generation rates were modified from 2011 trip-generation rates without any influence of NMO. The two scenarios are created to represent the two potential possibilities faced by stakeholders in their TDFM update process. A comparison of the three dependent variables using descriptive statistics is provided in Table 3.5. The reader will note that the trip rates are highest under Model B (NMO affects trip rates) while the base year trip rates are similar to trip rates for Model C. The research team employed a linear regression model for estimating the household trip- generation module. The base model specification was arrived at to closely match the activity-based model specified for the trip frequency model from the ARC travel demand model. The model estimation results from the base year and the two scenarios are presented in Table 3.6. The first column presents the variables considered in the analysis. The second, third, and fourth columns present the parameter estimates (with their t-statistics in parentheses) for the base model, the scenario where NMOs have an impact, and the scenario where NMOs have no impact. The overall structure of the update to the household trip-generation module is summarized in Figure 3.2. From the Table 3.6, some observations can be made. First, as expected, household socio- demographics, employment status, vehicle ownership, and income affect trips rates in the base year and two scenarios. Second, the base year model and the model where trips rates were not affected by NMO do not have any variable pertaining to NMO as affecting trip rates. Third, for the model where trip rates are affected by NMO, multiple variables are found to affect trip rates, including access to NMO for the household, NMO presence interacted with zero-vehicle house- holds, and NMO presence interacted with households with fewer cars than drivers. The results reflect potentially improved accessibility in the presence of NMOs for households with lower vehicle ownership. The parameters are intended to provide plausible impacts of NMO availability in the region. Finally, the reader should note that the models in columns three and four are estimated using simulated data and are provided for illustrative purposes only. Transportation practitioners seeking a TDFM update for trip rate will need to compile the trip-generation–rate data at the household level and develop specifications carefully. It is not necessary that the same variables be significant for every jurisdiction. The models and variable specifications need to be based on local, jurisdiction-specific data. Dependent variables Mean Std. dev. Min./Max. Number of HH trips for Model A 9.11 7.44 0/64 Number of HH trips for Model B 10.15 7.61 0/67 Number of HH trips for Model C 9.15 7.50 0/63 Table 3.5. Descriptive statistics of HH trips.

TDFM Updates 19   3.5 Use Case Example 3: Trip Mode Choice Model To assess the impact of NMOs on HH mode choice behavior, the research team developed two models using data sourced from a travel survey. Using an MNL approach, two mode choice models were estimated: (a) Model A, a base mode choice model with seven alternatives: (1) drive, (2) passenger, (3) transit, (4) walk, (5) bike, (6) park & ride/kiss & ride (PRKR), and (7) other; and (b) Model B, an updated model with nine alternatives considering availability of NMO in the region [the seven alternatives from base model plus (8) TNC and (9) shared micromobility]. For Model B, NMO alternatives were generated with the assumption that TNCs would draw travelers away from motorized alternatives such as driving, transit, and park & ride/kiss & ride, while shared micromobility would cause a decreased share of nonmotorized alternatives such as walking and biking. A comparison of the two scenarios is shown in Table 3.7. For these data, MNL model estimation results are provided in Tables 3.8 and 3.9. In the tables, the first column presents the variables considered, and the columns to the right present the Potential changes to the model componentsUpdates to considerUse Case Example 2 Household trip generation Change in independent variables Update linear regression model specification for traditional variables Include NMO variables Test NMO-specific variables to verify NMO impacts Variable Base Model A Model B NMOs Had an Impact Model C NMOs Did Not Have an Impact Constant 2.138 2.264 2.197(12.828) (13.381) (12.975) Number of workers 2.403 2.900 2.373(28.549) (34.456) (27.744) Number of retired persons 1.008 1.003 1.011(8.443) (8.399) (8.337) Number of homemakers 1.712 1.721 1.674(9.007) (9.049) (8.665) Number of students 2.058 2.071 2.030(19.440) (19.555) (18.871) Number of children 2.189 2.177 2.192(20.342) (20.222) (20.042) Zero-vehicle HHs 0.665 1.332 0.622(2.164) (3.760) (1.995) HHs with cars < drivers 0.778 0.893 0.836(4.080) (4.006) (4.311) HHs with cars > workers 0.890 0.892 0.934(6.680) (6.690) (6.897) HH income > $100,000 0.301 0.317 0.325(2.139) (2.252) (2.276) NMO availability indicator (urban HH indicator) — 0.844 — — (5.866) — NMO availability * zero-vehicle HHs — 1.233 —— (1.959) — NMO availability * HHs with cars < drivers — 1.368 —— (3.412) — Total records 10,278 Table 3.6. Household trip-generation module results. Figure 3.2. Updates to household trip-generation module.

20 New Mobility Options in Travel Demand Forecasting and Modeling: A Guide Mode Percent Share Base Model A (%) Model B Considering NMO Alternatives (%) Drive 64.7 61.7 Passenger 5.5 5.3 Transit 19.7 18.8 Bike 1.3 1.1 Walk 5.2 4.2 Park & ride/kiss & ride 3.4 3.4 Other 0.2 0.3 TNC — 4.0 Shared micromobility — 1.2 Table 3.7. Mode share by transportation alternative. Variable Base Model A Drive Passenger Transit Walk Bike PRKR Constant 3.623 0.770 3.667 4.577 1.187 1.043 (12.180) (2.352) (11.858) (14.573) (3.650) (2.698) Travel time –0.052 –0.052 –0.052 –0.052 –0.052 –0.052 –(16.272) –(16.272) –(16.272) –(16.272) –(16.272) –(16.272) Travel cost –0.346 — –0.346 — — –0.346 –(18.989) –(18.989) –(18.989) Origin distance from CBD — — — — — 0.080 (11.846) Destination distance from CBD 0.114 0.073 — — — — (17.531) (8.942) Male — –1.364 –0.295 –0.295 — –0.295 –(9.220) –(3.363) –(3.363) — –(3.363) Age ≤ 20 — 0.577 0.577 — 0.577 — (2.548) (2.548) (2.548) Age 21–40 — — 0.270 — — — (2.938) HHs with 1 person — –1.375 — — — — –(3.884) Zero-car households — — 1.107 1.107 1.107 — (5.062) (5.062) (5.062) HHs with 2 cars 0.967 — — — — — (11.173) Morning peak — — 0.347 — — 1.092 (3.362) (4.756) Table 3.8. Mode choice model result (base model). parameter estimates (and their t-statistics in parentheses) for the base model and the scenario where NMOs are available. For the model development, the “other” mode category was considered as the base category. Based on this base category, a positive (negative) sign of an independent variable for a mode choice alternative represents an increased (decreased) propensity for choosing the correspond- ing mode compared to the “other” mode choice category. It is important to note that all the independent variables (including trip characteristics, household demographics, and income attributes) used in the estimation process offered intuitive results, and the parameters estimated varied significantly from the base model (Model A) to Model B, where NMO is considered available. Finally, it is evident from the last three rows of Table 3.9 that, in regions where NMO is

TDFM Updates 21   Variable Model B, Considering NMO Alternatives Drive Passenger Transit Walk Bike PRKR TNC Micromobility Constant 3.664 0.565 3.477 3.744 0.237 2.281 –4.994 –3.604(12.654) (1.776) (11.897) (12.092) (0.619) (6.550) –(6.064) –(4.454) Travel time –0.054 –0.054 –0.054 –0.054 –0.054 –0.054 –0.054 –0.054–(17.144) –(17.144) –(17.144) –(17.144) –(17.144) –(17.144) –(17.144) –(17.144) Travel cost –0.362 — –0.362 — — –0.362 –0.362 –0.362–(20.945) –(20.945) –(20.945) –(20.945) –(20.945) Destination distance from CBD 0.076 0.042 — — — — — —(13.990) (5.511) Male 0.216 –1.144 — — 1.269 — 0.790 0.681(2.416) –(7.143) (4.177) (3.497) (2.329) Age ≤ 20 — — — — –1.919 –2.038 — —–(1.878) –(2.011) Age 21–40 — — 0.314 0.460 — — — —(3.419) (2.508) Households with 1 person — –1.204 — — — — — —–(3.409) Zero-car households — — 1.216 1.216 1.216 — –0.907 —(6.708) (6.708) (6.708) –(2.027) HHs with 2 cars 0.758 — — — — — — —(9.279) Morning peak — — 0.245 — — 0.935 0.245 —(2.573) (4.147) (2.573) NMO availability — — — — — — 5.637 5.637(7.830) (7.830) NMO availability * HHs car ≤ 1 — — — — — — 1.587 1.587(7.244) (7.244) NMO availability * high-income HHs — — — — — — 3.785 —(14.313) Table 3.9. Mode choice model result (considering NMO). available, the interaction of NMO availability with HHs having one car or fewer and high-income HHs is found to have higher utility for the NMO alternatives (TNC and shared micromobility). Transportation officials and stakeholders should carefully interpret the impact of these variables as their impact might vary across the mode choice alternatives depending on the regions and demographics. The overall structure of update for trip mode choice is summarized in Figure 3.3. 3.6 Summary This chapter outlined procedures for updating TDFM components. The data and scripts employed in the use case examples are provided in the appendix. The scripts can be used by practitioners to run the models on the data included and serve as templates for jurisdiction- specific TDFM updates. Potential changes to the model components Updates to consider Use Case Example 3 Trip mode choice Change in alternatives Incorporate additional modes as alternatives in the MNL model Change in independent variables Update linear regression model specification for traditional variables Include NMO variables Test NMO-specific variables to verify NMO impacts Figure 3.3. Updates for trip mode choice model.

Next: Chapter 4 - Summary and Guide Adoption »
New Mobility Options in Travel Demand Forecasting and Modeling: A Guide Get This Book
×
 New Mobility Options in Travel Demand Forecasting and Modeling: A Guide
Buy Paperback | $62.00
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Emerging transportation technologies and shared mobility services, or new mobility options (NMOs), are affecting travel behavior and demand. NMOs may include shared micromobility, transportation networking companies (TNCs), and connected and autonomous vehicles (CAVs). As NMOs grow in availability and use, transportation planners and decision-makers need to be able to understand how to harness positive and mitigate negative impacts.

NCHRP Research Report 1113: New Mobility Options in Travel Demand Forecasting and Modeling: A Guide, from TRB's National Cooperative Highway Research Program, provides travel demand modeling practitioners with ways to consider NMOs in travel demand forecasting models (TDFMs) - one of the primary tools available to understand potential impacts and future uncertainties.

Supplemental to the report are NCHRP Web-Only Document 399: Developing a Guide for New Mobility Options in Travel Demand Forecasting and Modeling; datasets of Use Case 1: Data, Code, and Tutorials for Household Vehicle Ownership Use Case; Use Case 2: Data, Code, and Tutorials for Household Trip Rates Use Case; Use Case 3: Data, Code, and Tutorials for Mode Choice Use Case; an Implementation of Research Findings and Products document; and a PowerPoint presentation of the research.

Any software included is offered as is, without warranty or promise of support of any kind either expressed or implied. Under no circumstance will the National Academy of Sciences or the Transportation Research Board (collectively “TRB”) be liable for any loss or damage caused by the installation or operation of this product. TRB makes no representation or warranty of any kind, expressed or implied, in fact or in law, including without limitation, the warranty of merchantability or the warranty of fitness for a particular purpose, and shall not in any case be liable for any consequential or special damages.

READ FREE ONLINE

  1. ×

    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?

    No Thanks Take a Tour »
  2. ×

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

    « Back Next »
  3. ×

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

    « Back Next »
  4. ×

    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.

    « Back Next »
  5. ×

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

    « Back Next »
  6. ×

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

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

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

    « Back Next »
Stay Connected!