Making Transit Sustainable And Equitable

Picture this: It’s Tuesday morning, and you’re planning to ride the train to work. Walking to the station takes 25 minutes, so you hop on the local bus. Today, though, the bus is delayed, and doesn’t reach the station in time to catch the train. You wait for the next one. You’re late for work.

If your boss is a stickler and you rely on public transit, a missed connection can be make or break. These are the kinds of problems that Samitha Samaranayake, a computer-scientist-turned-civil-engineer at Cornell University, has made it his mission to solve. He designs algorithms to help varied modes of mass transit work more seamlessly together—and help city planners make changes that benefit those who need them most.

Before Cornell, Samaranayake spent several years studying app-based ridesharing, including the potential of on-demand autonomous car fleets. In 2017, he co-authored an influential paper showing that companies like Uber and Lyft could reduce their contribution to urban congestion if cars were dispatched and shared efficiently. But he quickly became disillusioned with entirely car-centric solutions. “It’s convenient for people who can afford it,” he says, but when it comes to moving city-dwellers efficiently and accessibly, mass transit can’t be beat.

So Samaranayake began investigating how new technology can best be incorporated into city transit systems—and possibly solve some of their most-common pitfalls. Take the “last mile problem:” the challenge of transporting people from transit hubs in dense urban areas to the less-centralized places that they need to go—like their homes in far-out neighborhoods. If these connections aren’t quick and reliable, people may not use them. And if people aren’t using a neighborhood bus line or other last-mile service, says Samaranayake, a transit agency might cut it rather than run more buses, making the problem worse.

That’s where the technology developed by ride-sharing companies becomes useful, says Samaranayake. In recent years, he’s designed algorithms to integrate real-time data from public transit with the software used to dispatch on-demand vehicles. This could let transit authorities send cars to pick up groups of people, then deliver them to a commuter hub in time to make their connections.

This approach is known as “microtransit,” and after pandemic-related delays, a test project with King County Metro in Seattle launched earlier this year. It uses app-based rideshare vans to shuttle shift workers and others who live in the outskirts of the city to and from the regional rail line. Although it’s too early to measure success, Samaranayake has seen enthusiastic uptake from some commuters without many good alternatives.

That points toward his other goal: finding better ways to quantify how equitably transit resources are apportioned, so that city planners can ultimately design new systems that reach more people more efficiently. This social-justice element helps motivate Samaranayake to keep working on mass transit, even though funding has typically been more abundant for flashier technology like self-driving cars.

That could be changing: In recent years, Samaranayake and his collaborators have received nearly $5 million from the US Department of Energy and the National Science Foundation to pursue their vision. “Transit is not ‘cool’ from a research perspective,” Samaranayake admits. “But it’s the only path forward to a transportation system that is environmentally sustainable and equitable, in my view.”—M.G.

Publication: Javier Alonso-Mora, et al., On-demand high-capacity ride-sharing via dynamic trip-vehicle assignment, PNAS (2023). DOI: 10.1073/pnas.1611675114

Original Story Source: University of Arizona