Evaluating Demand Responsive Transit in Medium Sized Indian Cities through Agent Based Simulation

Abstract

Conventional fixed route buses struggle to satiate diverse, low density travel patterns pervasive in Tier-II Indian cities. The paper designs an agent based model (ABM) coupling synthetic populations with on demand minibus dispatch algorithms calibrated for Kanpur and Coimbatore. Demand responsive transit (DRT) scenarios door to door, corner to corner, and hybrid feeder to metro—are benchmarked against incumbent bus services over a 12hour simulation horizon. The hybrid feeder strategy exhibits the best tradeoff, boosting average occupancy from 24?% to 71?% while trimming passenger door to door travel time by 18?%. System level cost recovery improves by 35?% under a dynamic pricing regime. Sensitivity analyses highlight tipping points where DRT loses efficiency once ride request densities exceed 55?req?km²?h¹, indicating the need for adaptive service zoning.