Project

Validation of autonomous highway driving algorithm at Asta Zero

Period
1 August 2020–31 March 2021
Project manager
Johan Karlsson

Project B-0011 within SAFER's Open Research at AstaZero

Purpose and goals

The purpose of the project is to test, validate and demonstrate a control algorithm for overtaking scenarios in highway driving where a self-driving vehicle (AV) needs to pass a slower driving vehicle. Depending on the behavior of the slow vehicle, the AV needs to change its planned decision and trajectory (see video attachment for an example of such a solution in simulation). Especially we want to test the behavior at situations where the AV must abort a take-over and slow down and go back behind the slow vehicle.
So far these algorithms have only been validated in simulations. Moreover, another purpose is to gain insight into how testing and implementation can be done in the simulations step.

Results and effects

There are several expected results of the project:
- Insights of real-time aspects of the algorithms, such as computational time, robustness towards "real" disturbances, sensor noise and model errors.
- Material for scientific publication. The value of the theoretical results increases considerably by have been validated in test-driving. Also, the attention of the result increase.
- Attention to our research. Showing theoretical research results in real action gives credibility for future collaboration with industrial/international partners.
- New research questions. Based on the comparison of simulated and real test-driving, insights of the shortcomings will be gained, which is the start point for new research.

Test design and procedure 

The implementation of the project solution will be performed with the help of the Chalmers Revere laboratory located at Lindholmen in Gothenburg. The laboratory provides the hardware for autonomous tests, a Volvo truck fully equipped with sensors for autonomous driving and an additional car which can be used as a slow-moving leading vehicle. The laboratory also provides the software to interface between vehicles and the control algorithm via their program OpenDLV. In addition, AstaZero will provide an additional car that can be manually driven to act as a vehicle in the opposite lane.    

Short facts

Research area
Systems for accident prevention and AD
Financier(s)
Open Research at AstaZero Program (RISE, Chalmers)
Partners
Chalmers University of Technology
REVERE
Project type
Open research