AI recommender systems increasingly shape how urban residents choose to travel — but they are rarely designed with direct input from the communities they affect. This project uses a responsible-AI co-design methodology to develop a travel recommender that integrates multiple stakeholder perspectives: commuters, city authorities, transport operators, and environmental groups.
The work examines how algorithmic choices (objective function, fairness constraints, explainability) affect uptake and trust, and produces design blueprints transferable to other smart-city AI applications.
Presented at the 2025 International Conference on Information Technology for Social Good.