A Virtual Track Velocity Constraint for Integrity Enhancement in Automated Driving Systems

The integrity of the positioning information provided to Intelligent Transport Systems and Automated Driving Systems is crucial, as significant solution errors could escalate the risk of accidents. In fact, meeting Automotive Safety Integrity Level-D (ASIL-D) specifications demands decimeter-level accuracy and stringent Tolerable Hazard Rate criteria. However, the integrity of the velocity solution, often overlooked, must also be considered when Advances Driving Assistance Systems and other automated functions rely on Global Navigation Satellite Systems (GNSSs). Building on prior research, this work proposes to enhance the reliability of GNSS-derived velocity estimates thanks to the application of virtual track constraints. These constraints, originally applied in the railway domain, are adapted for automotive scenarios. The proposed system thus integrates GNSS with on-board sensors to align velocity estimates with the expected trajectory provided by a Digital Map of the road. Based on the track-constrained solution, we perform integrity monitoring by implementing the Solution Separation method in the parity space. The study demonstrates a reduction in the horizontal protection levels, emphasizing the effectiveness of virtual track constraints in mitigating errors on the estimated velocity.