I believe everyone agrees when I say that current public bus transport systems generally do not meet our transportation needs. Long travel and waiting times, buses which are overly full or completely empty, multiple uncoordinated transfers,… these are just a few examples of problems that public transport users and operators face.
Consequently, many people prefer to use their private car, causing all kinds of congestion issues. The million dollar question we asked ourselves at the beginning of my PhD trajectory was therefore:
“How can we organize public bus transport in a sustainable, flexible and efficient way?”
On-demand bus transport
What if buses would only drive when there is demand for transportation? What if all bus rides would be scheduled without any tedious transfers, regardless of your destination in a city? What if you would be guaranteed seating? We believe these characteristics would convince people to switch from private car to bus usage. Hence, let me introduce the concept of ‘on-demand bus transport’, the subject of my research.
“In this bus system, we, the passengers, would request our bus ride through a mobile application, just like ordering an Uber taxi.”
Consequently, central planning systems would route a fleet of minibuses, in a way that all passengers are picked up and dropped off as efficiently as possible. After issuing a request, every passenger is promised a latest arrival time, but the routes of the buses can change every minute. No bus would ever drive without a request for transportation, avoiding empty rides.
Plan your trip
Scheduling buses based on the demand for transportation is a difficult task and asks for powerful algorithms. That is because the planning system should be able to direct hundreds of buses and thousands of passengers simultaneously. In addition, it should be able to respond quickly when a new request is issued, as passengers do not want to wait hours for a system to respond.
Bus or taxi?
People have often asked me: what would be the difference compared to using Uber taxis? While taxis can stop anywhere, buses should stop at clearly signposted locations for safety reasons. Therefore, and also to more efficiently pool passenger requests, the algorithm should also smartly assign bus stops to passengers for both their origin and destination. Simply assigning a passenger to cross the street and take a bus instantly in the right direction, can make significant differences in passenger travel time.
By assigning bus stops to passengers, buses have to stop less frequently and have less detours to make. Of course, to maintain the comfort of passengers, the algorithm takes into account a maximum walking distance to the bus stops. Furthermore, because passengers are pooled into minibuses instead of taxis, the price of transportation is lower compared to Uber taxis.
Next steps
During my PhD research, we found that on-demand bus transport has the potential to replace traditional public bus transport (where all buses drive according to fixed routes and timetables). However, more research is needed before making a large-scale shift to on-demand bus transport.
We need to bring on-demand bus research from a theoretical to a real-life context and determine the key indicators of success. The latter will depend on the type of city, the demand for transportation, the existing public transport infrastructure, etc. Now, as an FBE postdoctoral researcher, I am looking forward to further contribute to improving public transport systems.
Lissa Melis is a post-doc researcher at the Department of Engineering Management, at the Faculty of Business and Economics.