Parallel shuttle bus service design for planned mass rapid transit shutdown: The Singapore experience

Document Type

Journal Article

Publication Date


Subject Area

mode - bus, mode - mass transit, mode - rail, place - asia, place - urban, planning - route design, planning - methods, infrastructure - maintainance


Parallel shuttle bus service, Bus route design, Bus terminal selection, Bus deployment, Bus berth assignment


19 MRT stations in Singapore experienced a two full-day shutdown as planned in order to upgrade the infrastructure of aging mass rapid transit (MRT) lines. A large bus fleet, running entirely parallel to MRT lines, was deployed to ferry passengers between affected stations, which however results in prolonged congestion and queue of buses near terminals. Rail transit systems in many metropolitan areas are threatened because of aging infrastructure. This study, motivated by the Singapore case, proposes a novel parallel shuttle bus service design (PSBSD) problem that addresses route design, terminal selection, berth allocation, and bus deployment to minimize inconvenience caused to passengers, which is seldom addressed in previous studies. We further develop a mixed-integer nonlinear programming (MINLP) model for the PSBSD problem. In order to solve the non-convex MINLP model optimally, we put forward an effective decomposition method, which comprises (i) the generation of candidate shuttle bus routes through a brute-force search for semi-route and a terminal selection procedure and (ii) the deployment of buses and assignment of terminal berths for selected shuttle bus routes by using an optimization-based approach. Numerical experiments were conducted using data sets from the MRT closure incident in Singapore. The results demonstrate that the decomposition method is capable of finding optimal solutions. Useful insights are also obtained from testing the models under different scenarios.


Permission to publish the abstract has been given by Elsevier, copyright remains with them.


Transportation Research Part C Home Page: