Motivation
Autonomous vehicles are being developed with the promise of improving road traffic safety, passengers’ conveniences as well as economic and environmental benefits. Autonomous functions that take control of the driving tasks off the human driver are heavily dependent on the system ability to perceive and understand the vehicle's surroundings through complex sensor systems. The sensor suite consists of radar, lidar, and video for situational awareness of the surroundings.
In the heart of autonomy, alongside artificial intelligence are these sensors which not only replace human ability to “see the road” but shall outperform human ability. Most of the sensors are actively emitting signals to use their reflections from the environment to understand the scene. The sensor data-processing algorithms produce understanding of the environment, enabling the autonomous function control system to adapt in response to prevailing conditions.
As connected and autonomous vehicles (CAVs) shall operate 24/7 in any weather and lightening conditions, the most robust and reliable sensor is radar. According to current vision each CAV will be equipped by at least a dozen of radars overlooking 360 degrees surrounding and seeing from a meter to hundreds of metres ahead and around.
- Co-existence and reliable performance of radars in CAVs are of critical importance for enabling reliable real-world L4 and beyond autonomy, in any environment.
- Multiple radars gives rise to co-existence challenges (mutual interference, environmental effects such as multipath, multi-bounce, etc.).
- Validated simulations are crucial to evaluate autonomous systems performance in complex scenarios and develop safety standards.
Objectives
- Improving road safety through the accelerated development of CAV technologies
- Supporting CAV sensors and algorithms design
- Providing effective testing and enhanced verification and validation of sensors
- Mutual interference suppression with cognitive sensors to modify parameters in real time
- Developing and validating optimal L4/L5 functions and generating recommendations to the wider CAV community