The AllDrive project aims to address the challenges of autonomous driving of heavy-duty vehicles in all-weather road conditions. The primary focus is to develop novel friction estimation together with road condition-aware motion planning and control for truck and trailer long-haulage vehicles and to demonstrate the system in real experiments. By the end of the project, the participants intend to have in place a framework for detecting road conditions a sufficient distance in front of and beneath the autonomous vehicle using appropriate sensors, sensor fusion, and off-board climate predictions. This will be used within modules of the autonomous vehicle to understand the situation of the vehicle and plan and control its motion. Throughout the project the improved capability will be evaluated on real vehicles in appropriate scenarios, such as braking, cornering, driving hilly terrain, and preventing trailer swinging or jack-knifing during maneuvers.
The project aims to make an important step toward all-weather autonomous driving, which will be a critical enabler for the safe introduction of autonomous heavy-duty vehicles and potentially also improve safety for manually driven vehicles.
Within the project, Klimator provides both predictive and detective technologies, RCD and AHEAD, to support the goal of enabling a fully functional friction estimation and autonomous vehicle system.
The overall goal is to conduct research and development that enables a fully functional friction estimation and autonomous vehicle system, which can drive safely under critical road conditions, such as wet or icy roads, thereby generating key competence that enables the future 24/7/365 operation of driverless vehicles.
The project's primary results include
1) Research and development of new friction estimation methods that can estimate friction under and in front of a heavy vehicle using a variety of sensors, signal processing, and self-learning methods.
2) Research and development of motion planning and control functions that enable the driving of heavy vehicles in adverse weather conditions.
3) Demonstrate the developed systems in several key scenarios.
4) Prepare friction estimation methods for autonomous and manually driven vehicles.
We have started the work to be able to deliver forecasts in the cloud to the system in the truck and created the first edition of the software that compiles forecasts from the cloud with detection in the vehicle. This is also run in the vehicle in real time and the impact is scanned by the self-driving system. We also installed LiDAR sensors and collected data to create the AHEAD camera algorithm
Where are we going next:
Releasing the first release of the AHEAD software that combines camera and LiDAR and developing the software that combined forecasting from the cloud with in-vehicle detection, to improve our knowledge of the road and driving conditions beyond the range of sensors.
Read more on the link below (SWEDISH) or contact us for more information.