GenAI for Sustainable Mobility Behavior Change: Building MUV's AI Driven Game Engine

Abstract

Achieving sustainability objectives necessitates a transformation in passenger mobility behaviour. This paper introduces a Generative AI framework developed by the MUV platform to address the enduring attitude-behaviour gap in sustainable urban mobility. Anchored in Self-Determination Theory, the system utilises behavioural, contextual, and qualitative data to create personalised competitions that enhance intrinsic motivation. The framework features a threelayer architecture and a virtual mobility trainer to support game dynamics, while federated learning and bias audits provide ethical safeguards. The framework is set to be piloted at ISPRA, with a mixed-methods evaluation planned to assess its impact on behaviour and inclusivity. This study presents a replicable model for responsible AI-driven behavioural change in alignment with the European Green Deal.