Imagine a self-driving car that doesn't panic when it suddenly rains or encounters a chaotic construction site. This work creates a digital 'safety net' that helps cars recognize when they are struggling and decide the safest way to react. It's like giving a vehicle the intuition to know when to slow down or pull over safely when things get too unpredictable.
Autonomous vehicles often fail or shut down when they hit 'Operational Design Domain' (ODD) limits, such as heavy snow, sensor failure, or chaotic urban traffic. This lack of resilience prevents the widespread adoption of higher automation levels.
A set of perception and decision-making algorithms and a self-assessment system for AVs. These were validated in simulation and real prototype vehicles.
If you are an AV System Integrator dealing with vehicles that fail in heavy rain or snow — this project developed perception and decision-making algorithms that allow the car to operate safely or perform a minimum risk manoeuvre. This reduces the risk of accidents during adverse weather.
If you are a Robotaxi Fleet Operator dealing with unpredictable pedestrians and cyclists in cities — this project developed VRU detection and interaction tools. This ensures the fleet can navigate complex urban environments without frequent system shutdowns.
If you are a Sensor Suite Provider dealing with high production costs for luxury AVs — this project developed a method to determine cost-efficient sensor suites. This helps you optimize the hardware mix for better market competitiveness.
Based on available project data, specific pricing is not provided, but the project explicitly aims to determine cost-efficient sensor suites to lower hardware expenses.
The project involves 8 industrial partners and tests algorithms in both prototype vehicles and simulations, suggesting a path toward industrial scale.
Based on available project data, licensing terms are not specified, but the project aims to promote results to standardisation bodies.
The project focuses on system integration of perception and decision-making algorithms into prototype vehicles and simulation environments.
The project period runs from 2022-09-01 to 2026-02-28, indicating that final results and validations will be available by early 2026.
The consortium is heavily weighted toward commercial application, with an industry ratio of 57% consisting of 8 industrial partners. With 14 partners across 9 countries, the project combines academic research from 3 universities and 3 research centers with practical industrial implementation, ensuring the results are grounded in market needs.
Contact the Research Center for Communication and Computer Systems (REC) in Greece
Contact us to connect with the EVENTS consortium for sensor suite optimization.
