Dial-a-bus services: Researchers investigate potential of algorithms
Whether you need to go to work, the train station, the doctor's, the grocery store, or you want to visit family or go on a trip – outside of cities, such trips are often made by private vehicle. This is because there is usually not enough demand for large buses that would run a fixed route to maintain an economical operation. This gap is filled by so-called dial-a-bus services such as the "VGI-Flexi" service offered by the Ingolstadt transport association (VGI), which is being scientifically supported by logistics expert Prof. Dr. Pirmin Fontaine from the Catholic University of Eichstätt-Ingolstadt (KU).
The research focus in this context is on options and mathematical tools for efficient and customer-friendly route planning of the dial-a-bus service. This is because, unlike conventional buses, on-call buses do not have a fixed schedule; the minibuses only travel when there are bookings. Fontaine holds the junior professorship of Operations Management at the KU Ingolstadt School of Management.
His subproject is part of a research consortium that also includes the Technische Hochschule Ingolstadt (THI) and the Artificial Intelligence Network Ingolstadt (AININ) in addition to the KU. The partners' common goal is to collaborate with VGI to strengthen local public transportation by using artificial intelligence. The overall "VGI newMIND" project is funded by the German Federal Ministry of Digital Affairs and Transport. The VGI's dial-a-bus pilot projects are among those funded.
In the case of VGI-Flexi, minibuses can be booked by app, phone, or online and then drive to a designated stop. In the transport area of Scheyern and Pfaffenhofen, 85 stops are available for this purpose. In turn, passengers can get off the bus at any address within the destination. The dial-a-bus service has been well received: Within the first month after the launch of the VGI-Flexi project in Scheyern, the service counted 1,494 passengers.
"A particular challenge when using demand-based bus systems lies in planning the bus routes. This has to be done within a narrow time slot between the latest possible incoming booking and the desired time for the pickup", explains Professor Fontaine. Customers of the VGI-Flexi program can book a trip between 60 minutes and 30 days in advance. If there is only one passenger, the route can take them directly to their destination, but if there are two, three or more passengers, the aim is to make as few detours as possible in order to travel quickly and economically. "The ideal situation is, of course, to pool customers with similar routes. But that is not always the case. This is where efficient algorithms offer potential for making trips as efficient as possible", says Fontaine. In another research focus, he is also looking into the potential of cargo bikes for urban transport of goods, which also raises the question of optimal tour design.
For dial-a-bus services, as Fontaine explains, it is possible to look at different scenarios determining which factors need to be adjusted in order have an influence on cost efficiency, maximum utilization of the buses or the service level for passengers. What advantage would there be, for example, if the system knew customer requests one day ahead, if possible? Would it be easier to plan the routes if flexible passengers agreed to be picked up ten minutes early or later? The results of such investigations could contribute to linking the attractiveness of such services to factors such as cost efficiency and a resource-friendly operation. Fontaine is convinced: “Public transport can make a major contribution to climate protection and a sustainable transformation of the transport sector.”
In addition, research projects from the USA on comparable services show that dial-a-bus services also have a social component. For passengers in Atlanta, Georgia, who cannot afford their own car, dial-a-bus services provide an opportunity to participate in urban mobility regardless. There, however, the researchers focused on the integration of on-demand services with regular bus routes that cover longer distances.
Prof. Dr. Pirmin Fontaine (Assistant Professorship of Operations Management; firstname.lastname@example.org; http://www.ku.de/wfi/om