Overview of ACUMEN Key Exploitable Results
Urban traffic management is becoming increasingly complex. Cities and transport operators must respond to congestion, incidents, shifting travel behaviour, and ambitious sustainability targets, all while managing networks that are more connected and multimodal than ever before. While digital technologies offer major opportunities, in practice they are often difficult to apply. Data is fragmented, tools are not always interoperable, and solutions do not always fit day-to-day operational needs.
ACUMEN set out to address this gap. The project focused on helping traffic and mobility practitioners make better use of data and AI to manage transport systems more efficiently, safely, and sustainably. Over the course of three years, we developed and validated a set of building blocks for multimodal traffic management. The building blocks cover a wide range of applications, such as enhanced traffic management strategies and novel incident detection frameworks.
- the challenge it addresses,
- what the solution does,
- and who can use it.
The building blocks originate from a research and innovation environment. As a result, they vary in maturity: from early-stage concepts to solutions close to operational use. While not all are ready for immediate deployment, they are designed to be adaptable and to support further development, integration, and scaling within real-world traffic management systems.
| Building block | Description |
| City Guidebook | A guide that helps cities integrate AI in traffic management responsibly and strategically. |
| Digital Twin | A modular platform that integrates models and data to simulate and analyse multimodal transport systems. |
| Dynamic perimeter control | A real-time strategy that regulates traffic inflow into congested areas using adaptive signal control. |
| Fleet rebalancing strategy | A learning-based method that repositions ride-hailing vehicles to better match future demand. |
| Generator framework for stress test scenarios | A framework that creates and analyses disruption scenarios to test system resilience. |
| Traffic incident detection method | A real-time method that detects unexpected traffic disruptions using standard data and ML techniques. |
| Traffic forecasting framework | A predictive model combining causal analysis and AI to forecast traffic conditions accurately and transparently. |
| Utility-based demand estimation method | A behaviourally grounded approach to estimating travel demand using aggregate data. |
| Incentive strategy framework | A framework for designing and evaluating incentive-based traffic management schemes. |
| Multi-objective trip pricing model | A pricing model that optimizes trips across multiple goals such as congestion, sustainability, and fairness. |
