Team design study · 2025
Privacy-preserving traffic management: a responsible AI design study
A conceptual traffic-management proposal examining privacy, stakeholder needs and human oversight.
Conceptual illustration; not an experimental result.
Design question
How could an AI-supported traffic-management system address mobility needs while limiting the collection and exposure of personal information? The report develops a proposal for an urban setting in Utrecht.
Approach
A literature-based design considers local processing, anonymisation and federated learning alongside stakeholder mapping, transparency and human oversight. Mock-ups make the proposed interactions and responsibilities explicit.
Contribution
The report connects architectural choices with ethical and social considerations, identifying trade-offs that should be addressed before implementation. It is a design and conceptual-analysis exercise.
Limitations
No deployed system or empirical traffic evaluation is reported. The proposal does not demonstrate measured privacy guarantees, traffic improvements or achieved legal compliance.
Related research
GPTSwarm: evolutionary graph optimisation