Improving BRT route design through code: The case of Bogotá’s BRT system, TransMilenio

Abstract

This study presents the methodology used by TRANSMILENIO S.A. –the company in charge of managing the BRT lines in Bogotá, Colombia– to renew its operational design. TRANSMILENIO S.A. is increasing its bus fleet by acquiring higher capacity bi-articulated buses. This process demanded a redefinition of the frequencies, stops, and type of bus for all routes in the system. Therefore, a line scheme was selected based on the number of trips between corridors. A non-linear optimization model was then used to establish an initial route design that could satisfy passenger demand in the morning rush hour. Five algorithms were implemented to adjust the initial route design to the actual restrictions, namely, fleet and bus capacity at stops. The algorithms modified the type of bus employed, the defined stops of routes, and eliminated low demand routes. Different alternatives were tested for the execution order of algorithms. The alternative to implement was selected based on compliance of the defined restrictions. Algorithms were iteratively run. Between iterations, a trip assignment model was used to adjust route frequencies based on demand. The execution order of algorithms, its impact on the level of service indicators, and compliance of restrictions are part of this study's discussion. Particular relevance and effects of each algorithm are also analyzed. Both methodology and results might be tested, adapted, and improved in further studies by other BRT systems as a reference or input for their own route design process.