Multimodal Data and Modeling
We need better data and tools for decision-making. Our research advances the collection, methodology, and analysis of multimodal data that supports professionals and researchers in understanding and predicting human travel behavior in order to optimize those systems for both the providers and users. These new models and tools examine the implications of changes to the system on a range of outcomes including equity, the environment, and health. Download the full literature review of NITC research in multimodal data and modeling here, or you can download our two-page summary here.
In a series of NITC Research Roadmaps, we surveyed a decade of contributions across six areas of transportation research funded by the National Institute for Transportation and Communities (NITC).
Nonmotorized Count Data
Counts of people walking and bicycling are important for monitoring trends, planning new infrastructure, and analyzing safety. NITC research has advanced the state of the practice by improving data quality and standardization, as well as by integrating new technologies. One study fused traditional count data with emerging data sources from GPS and mobile devices to estimate network-wide bicycle volumes, and another study is looking into a similar method for crowdsourcing pedestrian data. NITC researchers have explored methods to assess passively collected and crowdsourced data for accuracy, representation, completeness, and bias. Improved sensor technology has a role to play as well: one recent NITC study developed a new intelligent multimodal traffic monitoring device using low-cost mmWave radar.
To assess public transportation in a holistic way – in terms of rider experience, travel time reliability, accessibility, safety, efficiency, and environmental and economic impacts – practitioners need a phenomenal amount of data from a variety of sources. NITC researchers have made great strides in making it easier to gather and use transit-related data. One NITC study used high-resolution bus data to create more precise and statistically valid trip time models. Another compared bus stop-level data with demand-responsive parking data to show the effect nearby parking prices had on bus ridership. A third study used built environment factors – density, diversity, and walkability – to classify neighborhoods and found that transit-oriented development residents walk more, take transit more, and spend less on transportation than their counterparts in other types of neighborhoods.
Travel surveys provide information about travel patterns that is necessary for planning and policymaking. However, current survey methods do not adequately capture active transportation trips, and may not accurately represent marginalized populations. NITC researchers have significantly refined travel survey methods, improving accuracy and making the data more context-sensitive. One NITC study recommended the creation of pooled samples with comparable and consistent data from cities, which could be beneficial for understanding relationships between travel behavior and the built environment. Other NITC studies have looked at making data collection methods more inclusive of marginalized populations through strategies like increased transparency, protection of privacy, and more robust community engagement.
Applications of Multimodal Data: Equity
Despite recent improvements in household transportation survey methodologies, people of color and immigrant groups remain under sampled and understudied. A NITC study linked census population and housing data to a stratified random sample of households from the Oregon Household Activity Survey and found that the survey consistently overrepresented white households and underrepresented nonwhite households across the greater Portland area. The study developed a set of recommendations to improve the representation of diverse populations in travel surveys. Adoption of cashless fare systems has created barriers for lower-income transit riders. One NITC study explored the costs for agencies to maintain some cash options and found that simple approaches, such as cash collection onboard buses, can be quite cost effective at ensuring transit remains accessible and easy for all riders. Transit agencies currently lack guidelines for assessing the social equity impacts of replacing flat fare with distance-based fare structures. Another study found that shifting to a distance-based fare system may benefit low-income, elderly, and non-white populations; however, the effect is geographically uneven, and may be negative for members of these groups living on the urban fringe.
Applications of Multimodal Data: Economics
A national NITC study explored how investments in bicycle and pedestrian-focused street improvements can impact the economic vitality, business activities and neighborhood equity in surrounding areas across six cities. It found that, generally, street improvements yielded positive or non-significant impacts on business performance. Proximity to fixed guideway transit stations led to higher rents for commercial spaces and higher regional share of jobs closer to the transit station. One NITC study estimated development outcomes in response to transit and found market rent increases with respect to Fixed Guideway Transit (FGT) station proximity for all commercial types, but not for bus rapid transit (BRT). However, BRT systems are associated with positive development and job location outcomes. An earlier study revealed that BRT corridors gained new offices and multifamily apartments, and BRT station areas gained jobs in the manufacturing sector at a faster pace than the rest of the county.
Applications of Multimodal Data: Safety
Signal timing treatments can improve bicycle safety. A NITC study analyzed the operational impacts of various signal timing treatments on right-hook bicycle-vehicle conflicts. The research found that while a split leading bike interval treatment was useful in mitigating conflicts during the lead interval, the risk for bicyclists was then shifted to the stale green portion of the phase. The study also revealed that significant confusion was exhibited by both cyclists and drivers in a mixing zone, where bike and car traffic merged. The 2010 Highway Safety Manual provides methods for predicting the number of motor vehicle crashes on various facilities, but it includes a simplistic method for predicting the number of bicycle-related crashes. A NITC study developed the first bicycle-specific safety performance functions for segments in the U.S. and found that motor vehicle volume is a leading factor associated with more crashes between drivers and bicyclists. Bicyclist exposure, population density, and percent retail land use are also predictive.
Researchers and practitioners share common concerns about current data collection and processing strategies, especially a lack of standardization and validation. These limitations likely result in incomplete, inefficient, or even biased modeling strategies across all travel modes. Considering rapid changes in emerging data sources and technologies, as well as new travel modes, there is a clear need for greater scientific guidance and innovation on data collection, analysis, and modeling strategies.
What are the impacts and key findings of our research on how we can better collect, use, and model multimodal data? Learn more about some impact stories below.
Modeling tool supports transit agencies transition to electric buses while prioritizing environmental equity.
"The blocking piece is one of the more unique and helpful elements of this tool. We are making investments based on her recommendations, from the model and the tool, for five more high-powered chargers in our system.... You can optimize to a lot of different factors using her model. It's a really good tool in that you can use in multiple ways to make better business decisions for both your agency and the community."
-Manager of Systems Planning and Project Development, Utah Transit Authority
Learn more about Bi-objective Optimization for Battery Electric Bus Deployment Considering Cost and Environmental Equity, led by Xiaoyue Cathy Liu of University of Utah.
Photo courtesy of TriMet
Bike-Ped Portal offers a centralized standard count database for non-motorized data nationwide.
Another research project funded by NITC, led by Nathan McNeil of PSU, offers a method for monitoring the quality of this bike-ped count data. "There has been an effort to collect more bike-ped count data in recent years, but it hasn't been consistent in terms of what's being collected and how it's stored. If the data aren't in a uniform format, or aren't stored in a location where they can be easily accessed in bulk, then doing a deep scan of the data would be a challenge," McNeil said.
Researchers at the Mineta Transportation Institute of the San Jose University used BikePed Portal data to examine the consistency between crowdsourced and traditionally collected count data to obtain more accurate bicyclist and pedestrian counts, which is critical to better designing active transportation-related facilities and empowering people who walk and cycle.
Learn more about Biking and Walking Quality Counts: Using “BikePed Portal” Counts to Develop Data Quality Checks and access the BikePed Portal here.
Visual analysis tools highlight the usefulness and value of GPS trajectory data.
We are also using GPS data in the analysis of the Istanbul BRT line with fully GPS-equipped buses and also big data of electronic fare card records to investigate the passengers' trip movements between origin and destination stations during peak hours and the day. In this respect, the final products of your research are useful especially in developing our visual presentations of the outcomes.
-Associate Professor, Yildiz Technical University
Learn more about Visual Exploration of Utah Trajectory Data and their Applications in Transportation, led by Nikola Markovic of University of Utah.
New bike count models combine traditional counters and emerging GPS data
“At ODOT we just adopted "Bicycle Miles Traveled" as a new key performance measure, and we need a way to measure it, so this project very much helps to fill the gap on how we're going to do that. This research used cutting-edge data fusion techniques that could lay the groundwork for how transportation agencies like ODOT monitor bicycle activity across the system.”
-Josh Roll, Research Analyst & Data Scientist at the Oregon Department of Transportation
Learn more about Exploring Data Fusion Techniques to Derive Bicycle Volumes on a Network, led by Sirisha Kothuri of PSU.