PreMAT - Predictive Monitoring for Autonomous Trucks

Purpose and goal:

PreMAT seeks to lower the thresholds for introduction of autonomous driving systems onto the market by providing an extra layer of safety necessary to perform early testing in relevant environments. The project aims to answer the question how to guarantee that an automated vehicle never exits a predefined physical space, and thereby guarantee the safety (w.r.t. the vehicle) outside of it.

Expected results and effects:

The effects, once introduced to the market (outside project time frame) is an increase in test safety, an increase in possible test scope as well as a reduced toll on supervisory personnel. The method is expected to be applicable to other areas where early adoption of AVs is possible such as parking areas, open pits, or mines however the first point of contact is the AstaZero proving ground.

Approach and implementation:

During the entire project, data will be collected to be used for extensive system testing and pattern identification. There are two parallel tracks in the project: one makes use of the data to prototype a data clustering-based supervision method, the other develops a deterministic solution and makes use of the data to show reliability of the proposed solution. The positioning accuracy of the Axis and Viscando systems for real-time applications will also be evaluated.