Making Self-Driving Car Sensors Work Reliably in Rain, Fog, and Snow

Imagine your car's cameras and sensors are like your eyes — they work great on a sunny day but struggle in heavy rain or fog. That's exactly the problem self-driving cars face: their sensors fail when conditions get tough, which is precisely when you need them most. RobustSENSE built a system that lets sensors monitor their own reliability and automatically adapt when conditions worsen, so the car always knows how much it can trust what it "sees." Think of it as giving the car the common sense to slow down and be more careful when visibility drops, rather than just shutting off entirely.

What needed solving

Autonomous vehicles and advanced driver assistance systems rely on sensors (cameras, radar, lidar) that degrade or fail in bad weather — rain, fog, snow, glare. When these sensors lose accuracy, the system either makes dangerous decisions or shuts down entirely, leaving drivers stranded or causing accidents. Companies developing autonomous vehicles need a way to keep their systems operating safely across all environmental conditions, not just on sunny days.

What was built

The project built a self-diagnosing sensor system that monitors its own reliability at every level — from individual sensors up to the decision-making layer — and automatically adapts its behavior when performance drops. A publicly presentable demonstrator with the full integrated system was delivered, along with 6 total deliverables covering the sensing, decision, and planning pipeline.

Who needs this

Who can put this to work

Quick answers

Who built it

This is a heavily industry-driven consortium with 60% of partners (9 out of 15) coming from industry, coordinated by Mercedes-Benz Group AG — one of the world's largest automakers. The presence of 3 SMEs alongside major industry players across 5 European countries (Austria, Germany, Spain, Finland, Italy) signals that the technology was developed with real commercial pressure and supply-chain relevance, not just academic curiosity. The single university partner and 5 research organizations provided the scientific backbone, but the consortium's center of gravity is clearly industrial, which increases the likelihood that results are production-relevant.

• MERCEDES-BENZ GROUP AGCoordinator · DE
• EUROPEAN CENTER FOR INFORMATION AND COMMUNICATION TECHNOLOGIES GMBHparticipant · DE
• ROBERT BOSCH GMBHparticipant · DE
• FZI FORSCHUNGSZENTRUM INFORMATIKparticipant · DE
• CENTRO RICERCHE FIAT SCPAparticipant · IT
• SICK AGparticipant · DE
• TEKNOLOGIAN TUTKIMUSKESKUS VTT OYparticipant · FI
• UNIVERSITAET ULMparticipant · DE
• AVL DEUTSCHLAND GMBHparticipant · DE
• FUNDACION PARA LA PROMOCION DE LA INNOVACION INVESTIGACION Y DESARROLLO TECNOLOGICO EN LA INDUSTRIA DE AUTOMOCION DE GALICIAparticipant · ES
• MODULIGHT OYJparticipant · FI
• AVL LIST GMBHparticipant · AT