Human-Machine collaborative decision-making approach to scheduling customized buses with flexible departure times

• Published in: Transportation research. Part A, Policy and practice Vol. 187; p. 104184
• Main Authors: Liu, Tao, You, Hailin, Gkiotsalitis, Konstantinos, Cats, Oded
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.09.2024
• Subjects: Customized bus; Flexible departure time; Human–machine collaborative decision-making; Public transport; Vehicle scheduling; Vehicle scheduling; Public transport; Customized bus; Flexible departure time; Human–machine collaborative decision-making
• ISSN: 0965-8564
• DOI: 10.1016/j.tra.2024.104184

Public transport agencies need to leverage on emerging technologies to remain competitive in a mobility landscape that is increasingly subject to disruptive mobility services ranging from ride-hailing to shared micro-mobility. Customized bus (CB) is an innovative transit system that provides advanced, personalized, and flexible demand-responsive transit service by using digital travel platforms. One of the challenging tasks in planning and operating a CB system is to efficiently and practically schedule a set of CB vehicles while meeting passengers’ personalized travel demand. Previous studies assume that CB passengers’ preferred pickup or delivery time is within a pre-defined hard time window, which is fixed and cannot change. However, some recent studies show that introducing soft flexible time windows can further reduce operational costs. Considering soft flexible time windows, this study first proposes a nearest neighbour-based passenger-to-vehicle assignment algorithm to assign CB passengers to vehicle trips and generate the required vehicle service trips. Then, a novel bi-objective integer programming model is proposed to optimize CB operation cost (measured by fleet size) and level of service (measured by passenger departure time deviation penalty cost). Model reformulations are conducted to make the bi-objective model solvable by using commercial optimization solvers, together with a deficit function-based graphical vehicle scheduling technique. A novel two-stage human–machine collaborative optimization methodology, which makes use of both machine intelligence and human intelligence to collaboratively solve the problem, is developed to generate more practical Pareto-optimal CB scheduling results. Computation results of a real-world CB system demonstrate the effectiveness and advantages of the proposed optimization model and solution methodology.