OptiRoutS
20014
From 2022 to 2024
The OptiRoutS project developed a predictive model that combines classical traffic models and AI to calculate real-time traffic situations and external costs to guide users along socially desired routes. TML was responsible for building hybrid traffic models and translating traffic situations into real-time external costs.
This project used a combination of classical traffic models and artificial intelligence to develop a predictive model for real-time traffic situations. These traffic situations formed the basis for the calculation of social external costs (safety, environment, noise, etc.) used in a route planner. In the process, users can be encouraged to follow a socially desirable route.
OptiRoutS aimed to realise a new in-car traffic management ecosystem that proactively contributes to the achievement of a wide range of mobility goals. It aimed to replace routing systems based on expensive variable information signs. Thus, OptiRoutS aimed to better align road authorities' goals regarding end-users' routing and their broader sustainable objectives. This effect is achieved by guiding individual users along more socially desirable routes.
The final product of this project comprises several sub-elements:
- A hybrid predictive model for real-time traffic conditions.
- A methodology for converting traffic states into external costs.
- A route planner that guides users along socially desirable routes based on these external costs.
Within this project, we worked with five work packages:
- WP1 examined the case studies and architecture and derived the critical performance indicators (KPIs) for pro-social routing and the stakeholders involved. These findings led to the definition of a new sustainable management model to align the public and private ecosystem.
- In WP2, we designed hybrid models for traffic situation estimation for vehicular traffic. These fed predictions and optimisation of network design in WP3.
- WP3 then created market value from WP2's model results by translating them into external costs that quantify the previous KPIs that stakeholders want to optimise. WP2 also explored the impact at both system and network level and of network design decisions on these KPIs.
- WP4 explored how we can encourage users to follow socially desirable routes derived from external costs through a system of altruistic rewards. We also investigated how to improve this use of both optimal and compromise routes.
- Finally, WP5 collected the results of WP1 to 4 in a proof of concept with end-users.
TML was responsible for building out the hybrid traffic models and translating traffic situations into real-time external costs.
