A Duluth Transit Authority bus on a rural road in winter. A bike is secured to the bike rack at the front of the bus.

Enhancing Public Transit with Autonomous Mobility Service to Cover the “Last Mile”

The benefits of public transportation, including reducing individual vehicle use and traffic congestion, aren’t always optimized beyond densely populated urban settings. While transit may extend to more suburban or rural areas, land development patterns can make it challenging for people to reach their destinations by transit alone. Pairing transit with an autonomous mobility service is a potential solution. Researchers assessed transportation needs in a suburban shopping area and developed a system designed to deliver passengers from the end of a transit line to their ultimate destination. Simulations of the autonomous mobility service revealed its potential to increase transit efficiency and reduce individual vehicle use.

Public transportation can provide a safe, affordable, convenient and environmentally friendly transportation alternative to single-occupancy vehicles, but coverage is inherently limited. Distances from a transit termination point to a passenger’s final destination can be lengthy or involve unsafe walking conditions. The “transit last-mile access” issue can be prevalent in suburban or rural areas where land use patterns are oriented toward individual vehicles. 

An autonomous mobility-on-demand (AMoD) system has a fleet of autonomous vehicles that picks up passengers and transports them to a destination. Operating within a limited geographical area, AMoD uses a control system to manage the fleet and meet transportation demand by assigning passengers to vehicles and routing them to continuously balance the fleet’s geographical distribution. AMoD is being tested in Minnesota and elsewhere in the country, though current Minnesota uses do not include last-mile service. 

Transportation agencies seek to reduce vehicle miles traveled for multiple reasons. Given AMoD’s potential responsiveness to travel demand, fleet repositioning and route flexibility, LRRB wanted to explore the system’s viability in an area with significant transit last-mile access issues.

What Did We Do?

After a review of past research on AMoD services, including passenger preferences, planning and operational considerations, the project’s Technical Advisory Panel helped to identify a case study area. 

The Miller Hill Mall (MHM) in Duluth is a large commercial area with multiple shopping plazas. Historically, the area has presented a challenge for pedestrians due to building setbacks; absence of sidewalks; and multilane, high-speed roadways. 

“Effective public transit is challenging in areas of car-oriented land use and development. But with further testing and support, an autonomous mobility-on-demand system could be one solution,” said Chris Belden, planning and grants director, Duluth Transit Authority.

Although the area has public transit service, not all surrounding areas in the region can easily access it. Additionally, moving efficiently and safely among establishments without a vehicle can be impossible for pedestrians. 

Data on transit service in the Duluth area and passenger use illustrated accessibility and service gaps. A travel time analysis from regional locations to and around the MHM area explored different travel modes, times of day and safety of walking between destinations in the area.

To design and evaluate an AMoD service, researchers used mathematical modeling and algorithms to determine the necessary fleet size, geographic balance and routing to meet travel demand at different times. 

The system, integrated with the existing transit network, included a transit hub where passengers would transfer from a fixed route to an AMoD vehicle. Researchers simulated the system’s operation for an existing travel demand, measuring travel time, waiting periods and other relevant factors.

What Did We Learn?

The transit analysis of the Duluth area and MHM indicated that it could take up to four times longer from some parts of the region to access MHM by transit than by individual vehicle. 

Once in the area, people may have to walk 20 to 30 minutes between stores and services, and may have to cross busy highways or streets. 

The simulation of the AMoD system that researchers developed illustrated its potential to enhance transit efficiency, increase transit demand and reduce individual vehicle use. Larger fleet sizes produced more efficient service delivery such as reduced waiting times. 

What’s Next?

Various research efforts around the country and some limited implementation illustrate AMoD as an emerging technology. A demonstration project could eventually test an AMoD system for last-mile transit, though vehicles would initially include a driver. 

Potential topics to explore in the future include passenger attitudes and preferences, electric fleets and ride-sharing, and policy and legal issues.

More Information

Leave a Comment