Compared to rail systems, bus networks can be altered at relatively low cost to accommodate changes in demand. Bus network revisions include large-scale overhauls, such as recent redesigns in Houston and Columbus, as well as incremental approaches to bus network change, such as route additions, deletions, and realignments. To make these changes, planners need to understand how the current system is used and where there is potential for improvement.
In recent years, contactless smart card-based automatic fare collection (AFC) systems have become increasingly prevalent. Some agencies, including Utah Transit Authority (UTA) and Chicago Transit Authority (CTA), also accept payment via contactless credit and debit cards. These technologies allow cards to be tracked over time. They provide comprehensive ridership data, enabling passenger-centric evaluation and impact analysis. In contrast to survey data, this information is collected for all times of day and days of the year. Analysts can evaluate the behavior of a panel of individuals over time, including before and after a redesign.
Agencies like Boston’s MBTA, Washington Metropolitan Area Transit Agency (WMATA), and New York City Transit (NYCT) have applied methods to AFC and automated vehicle location (AVL) data to infer complete passenger journeys within their transit system. This origin-destination level data allows for analysis of passenger transfer behavior and path choice. In the future, agencies are likely to draw on new data sources from other modes including driving, Transportation Network Companies (TNCs), and bike sharing.
Data and Network Design in Practice
For Transport for London, Cecilia Viggiano of EDR Group developed a framework for applying AFC data to efficiently identify opportunities for valuable bus network revisions. The framework defines metrics to identify origin-destination (OD) pairs with significant potential for improvement, based on complete journeys inferred from AFC and AVL data. Improvable OD pairs are characterized by long travel times and distances relative to driving, and multi-stage journeys. Based on expected demand and benefits, this methodology groups improvable OD pairs into corridors that are candidates for new bus service.
A key component of bus corridors is the density of demand to support bus service. In the London network, 11 corridors had sufficient demand and expected benefits to justify new service. Interestingly, these corridors would serve less than 10% of the demand from the improvable OD pairs. The remaining improvable OD pairs were spread across London with a lack of density to support new bus service. Transit agencies that want to serve lower demand density trips may find solutions in other modes, such as microtransit or on-demand services.
Can Data Solve All Our Planning Problems?
Discussions about data-driven planning often ask whether data can replace planner knowledge and experience. The answer is no. The framework developed for Transport for London uses planner-defined parameters, which can be specified based on conditions in the network and planning priorities. Planner expertise and public feedback are important inputs to the decision-making process. Data-driven analysis should be treated as a tool to help reveal information planners wouldn’t otherwise know and provide evidence and justification for decisions. Because these methods can be automated, they can enable planners to regularly monitor networks and consider more frequent network and service changes, which can help transit adapt in the changing mobility environment.
Interested in more innovative uses of new transportation data? EDR Group used new vehicle speed data to evaluate the effectiveness of different incident management systems on rural highways. Urban streets may also soon be candidates for these types of analysis as more and more powerful tools become available.