Belleville Transit recently piloted an app to enable riders to hail on-demand rides on a city bus—becoming the first city in North America to try this innovation. Bilal Farooq, Assistant Professor in Transportation Engineering at Ryerson University, studied the pilot data with his research team. This report on their analysis, prepared by Ryerson CBI, presents the findings in summary.
In 2018, the Ontario city of Belleville, population approximately 50,000, launched a trial run of a new app-based solution to enable passengers to hail and schedule transit trips on the city’s late-night Route 11 bus, from 9 PM to midnight on weekdays and 7:30 PM to midnight on weekends. The pilot was an experiment to transform this traditional, fixed bus transit route into an on-demand transit (ODT) service. The technology that made this trial possible was developed by the company Pantonium, and Belleville is the first city in North America to test it.
During the pilot period, an app created for the service enabled users to book their bus trips on RT 11, and specify desired timing, pick-up and drop-off points, number of riders to be picked up and wait-time tolerances. (A call-in service was also provided as an option.) Users could choose to get on and off the bus at traditional stops, or create their own stops anywhere within the RT 11 zone. The bus would change its route in real time according to the pick-up and drop-off demand points. In essence, the pilot brought the conventional bus service closer to a ride-hailing experience.
Following the nine-month pilot project, in early in 2019, researchers from the Laboratory of Innovations in Transportation (LiTrans) at Ryerson University analyzed data collected by Pantonium. To unpack the results of the pilot, the researchers at LiTrans evaluated the spatio-temporal demand, supply, level of service, origin and destination patterns of Belleville’s ODT users. They were particularly interested in using this unique dataset to determine if an ODT system could successfully address the main challenge of conventional public transit in low-density areas: how to maintain a reasonably high frequency and low operating cost per trip at off-peak hours.
The researchers mined the data behind 14,132 ODT bus trips for spatial, temporal and origin-destination patterns. After sorting active users of Belleville’s on-demand transit app (1,420 people) into groups by frequency of usage, they reviewed users’ travel habits.
Since users were able to specify ad-hoc locations for pick-up and drop-off, and alter the conventional fixed route of RT 11, the data revealed the origins and destination locations most desired by riders. Most of the top users had the same origin and destination patterns. They also had similar timings for most of their trips, when the team examined temporal trip distribution.
Key FindingsThe results were promising. The Belleville pilot has successfully captured many of the benefits of ODT—including higher efficiency, increased coverage, and greater convenience and reliability for riders.
The city’s ODT app was downloaded 2,074 times during the pilot, showing significant uptake. The on-demand service saw an average of 1,570 monthly trips. Compared to Jan-Mar 2018 ridership for the fixed bus route, the on-demand bus route in Jan-Mar 2019 saw a sizeable increase in ridership.
From a calculation of the most popular stops, the research team extrapolated travel patterns and behaviour. Overall, riders used ODT bus service for various purposes including work, shopping, education, health and leisure. Some of the top uses included commuters travelling to the main intercity bus terminal and people going to and from the local community centre.
The #1 pick-up spot (21.92% frequency) was the Walmart Belleville Supercentre—an area with 24 stores and warehouses. Both zonal and individual level data strongly suggest that the primary use of the ODT service comes from people working within the Supercentre travelling home following their shifts on weekdays. Further demographic analysis confirmed that the majority people living in the zones where trips were concentrated belong to the middle and working class.
Regarding responsiveness and reliability, the ODT system performed well. Of the total trip requests made by app users, call-in riders, and walk-on riders, 98% were successfully assigned. Wait times were satisfactory; a majority of wait times fell within the stated tolerances of app users. (Other riders were not able to specify wait time tolerances.) With its dynamic routing, the ODT service served more pick-up and drop-off locations (311) compared to the 93 fixed stops of RT 11, indicating an increase in users’ convenience in terms of lesser distances for access and egress. Whereas the conventional RT 11 served the periphery of Belleville’s central area, a high number of origins and destinations (ODs) of the ODT service were concentrated in the city centre. (See figure of OD patterns on previous page.)
This data analysis is helpful to Belleville Transit, as it evaluates the pilot and the broader potential of ODT to optimize its network of 10 bus routes and fleet of 15 buses. Other cities may also learn from Belleville’s experiment as they grapple with similar challenges of high operating costs and low ridership along low-density routes. It also represents an option to integrate the ride-hailing experience into public transit.
Ultimately, if cities such as Belleville can improve mobility for riders by adapting existing infrastructure to provide a higher level of service, they can improve the competitiveness of public transit against ride-hailing services dependent on private automobiles, and potentially reduce the number of overall vehicle miles driven.
Research performed by Shadi Djavadian, Bilal Farooq and Irum Sanaullah, Department of Civil Engineering, Ryerson University. For more information contact bilal.farooq@ryerson.ca.
