The most common way to address reliability is to hold buses at certain critical stops aiming at keeping similar headways ahead and behind every bus. However, Delgado et al (2012) shows that if holding is implemented in a myopic way (as sapos always did), it can end up over-holding buses and therefore damaging waiting and travel times excessively. Many other control strategies have been proposed: station skipping, boarding limits, overpassing-and-expressing, transit signal priority, etc.
In this chapter we present and discuss holding control strategies that use a complete knowledge of the system based on our Holding Real-Time optimization model (HRT) presented in Delgado et al (2012). First, we present a classification of control strategies according to two different criteria. Second, we introduce and explain the HRT control strategy, the data requirements and technology necessary for a complete implementation. Third, two case studies are presented in order to highlight the benefits of this strategy. The first case study is a simulation experiment where we compare the proposed strategy against two benchmarks. In the second case study, we present the results of a pilot program run in the Transantiago system in Santiago, Chile. Finally, we summarize the main results by evaluating the influence on the level of service to users and the cost to operators for all three strategies and discuss the main challenges in the implementation of these kinds of strategies in real operations.