State-Based Behavioral Modeling Framework for Electric Taxis: Extending from Traditional Fleets to Unmanned Fleets

As electric and autonomous mobility advances, understanding electric taxi (ET) behavior is vital for urban planning. Traditional ETs (TETs) are limited by driver rest and subjective decisions, reducing service consistency. Unmanned ETs (UETs) can overcome these constraints through automation, yet their large-scale urban impacts remain underexplored. This study develops a state-based simulation framework modeling TET and UET operations are modeled across four states (search, work, charge, rest), with transitions determined by time, energy, and spatial demand. Using Shenzhen GPS data, the TET model is validated for realism, then adapted for UETs by removing rest constraints and adding adaptive exploration and charging. Simulations show UETs achieve higher, more stable pick-up rates and more balanced, distributed charging, highlighting their potential to enhance service efficiency and infrastructure use.