SHARED ON-DEMAND AUTONOMOUS BYCICLES IN DONOSTIA
The development of autonomous driving systems has become a major focus of technological innovation and research in recent years. However, autonomous driving technologies are typically conceived for conventional motor vehicles. This raises a compelling question: what if we redirect these innovations toward ultralight and city-friendly vehicles, such as bicycles?
We have explored this idea by modelling the existing DBizi fleet in San Sebastian under a hypothetical autonomous scenario. The project assesses the impact of deploying a shared, on-demand bicycle system in Donostia-San Sebastián. Using Agent-Based Modeling (ABM), the platform simulates the behavior of these emerging last-mile mobility solutions in a mid-sized urban environment.
This tool serves as a valuable aid in decision-making processes related to the planning and management of next-generation shared mobility services in urban environments. To make the tool more accessible to relevant stakeholders—and to support its use in participatory planning processes—it has been implemented on a physical, interactive table, as shown below.
Physical Interface
Digital Interface
It allows users—including policymakers, mobility providers, and urban planners—to compare the performance of the traditional dock-based bike-sharing system with an autonomous, on-demand alternative. This simulation allows us to experiment with key operational parameters such as fleet size, battery capacity, and cruising speed. This flexibility supports scenario analysis and policy experimentation, helping the transition to the future autonomous scenario.
The project serves as a practical use case within the broader research initiative MoBiScope, which focuses on innovative mobility systems. More information about MoBiScope can be found here. The data used for this project was provided by Dbizi, the city’s official bike-sharing system operator. Their anonymized usage data was instrumental in developing the ABM and conducting simulations.
Traditional Scenario
Autonomous Scenario
The model provides insights into the expected usage patterns of the fleet, including the number of bikes that are charging, idle, or in service throughout the day, as well as average user waiting times. It also reports the number of completed (satisfied) trips in comparison to unfulfilled requests. These evaluation criteria help policymakers compare the two scenarios and, within each scenario, assess the impact of changes to key system variables.
