07 May 2025

The ACUMEN project is developing a governance and regulatory framework, as well as a decision support tool, to guide policy makers in integrating artificial intelligence (AI) and machine learning (ML) into transport management systems. To ensure practical relevance, ACUMEN conducted four workshops with its pilot cities – Luxembourg, Athens, Helsinki and Amsterdam – bringing together national and local traffic management stakeholders.

These workshops provided an insight into existing traffic management systems and facilitated the exchange of best practices. Feedback from participants will be used to refine the governance framework and decision-making tool to ensure its applicability in real transport systems. The final tool will be shared with stakeholders to demonstrate the benefits of AI-based traffic management and encourage cities to adopt AI-based solutions for long-term mobility improvements.

The Amsterdam pilot focuses on managing network disruptions, such as tunnel closures, through joint multimodal traffic management. Using historical data and simulations, the virtual pilot tests multimodal strategies – such as improving the distribution of users between modes and reducing delays and emissions. By involving stakeholders and aligning with the city’s long-term mobility vision, the pilot supports more resilient, data-driven traffic management and informs future policy decisions.

Interviews were conducted with representatives from the city of Amsterdam to understand how the city currently handles traffic management and where AI could fit in.

Amsterdam’s traffic operations are currently managed by a rule-based system that triggers pre-defined scenarios based on traffic conditions. While this approach works in routine situations, it becomes less effective when dealing with broader objectives such as sustainability or air quality. Developing rules that take into account multiple variables quickly becomes complex and difficult to manage.

A lack of detailed data, particularly on pedestrians and cyclists, adds to the challenge. Much of the available data comes from private providers, which limits independence and consistency in planning. While there’s growing interest in using artificial intelligence, its current use remains limited. There are also concerns about transparency – any future system will need to provide clear and understandable results, particularly for public accountability.

Amsterdam’s future approach involves gradually shifting from rule-based systems to AI-supported tools. Rather than replacing traffic operators, AI is expected to support them by analysing trends, forecasting outcomes, and proposing interventions. This approach depends on having reliable data and clearly defined transport objectives that AI can work towards.

The governance and policy environment is still taking shape. The current regulatory framework does not fully address AI use in traffic systems, and the city is working on translating its political goals into measurable indicators that AI tools can support.

Several actions have been identified to move forward:

• Develop transport performance indicators that reflect the city’s priorities.
• Ensure that AI tools remain transparent and provide understandable recommendations.
• Improve access to real-time data for all modes, especially for pedestrians and cyclists.
• Strengthen cooperation with private data providers and define clear data sharing agreements.
• Build open, modular systems that allow flexibility and avoid reliance on a single vendor.