Designing and planning sustainable customized bus service for departing attendees of planned special events: A two-phase methodology integrating data-driven and demand-responsive

Planned special events (PSE), such as concerts and sporting events, will seriously affect the normal operations of the transportation system, especially at the egress stage. The huge taxi and ride-hailing trips of departing attendees is one of the prime reasons for traffic congestion, meanwhile, insufficient vehicles and traffic congestion around the venue also lead to excessive passenger waiting time. Customized bus (CB) service is an emerging green sustainable transportation mode, and its characteristics are suitable for solving corresponding problems. This paper designs a novel CB service mode and constructs a corresponding two-phase planning methodology integrating data-driven and demand-responsive. The first phase includes a potential demand identification model based on graph semi-supervised learning and an integrated algorithm to deploy CB stops. The second phase is CB route planning with heterogeneous vehicles based on attendee choices. A concert held in Tianjin, China was selected as case study to evaluate the designed CB service. The results show that compared with the corresponding taxis and ride-hailing trips, the generated CB services can save 37.71% of the average travel time, 36.56% of the expenses per passenger and 89.58% of the spatial road occupancy, which can greatly alleviate traffic congestion, reduce carbon emissions and improve attendee travel.