Kontrol KoPilot is an independent, secondary safety system within the vehicle.
With Kontrol KoPilot our customers can verify if complex assistance systems are following traffic regulations.
- Real time validation of regulatory compliance
- Safe and certified vehicles based on independent technology
- True fail operational
- Sensor senses its surroundings
- Perception and Localization: understanding the world
- Behavior/Motion planning: Calculating trajectory and Behavior
- Vehicle Control: Follows the trajectory, and controls the UAV/car
Input Parameters:
- State of the vehicle
- Object list from the perception
- Other information: Map, Location, Perception Quality, etc.
-
Written Rules1
-
2Translation into formal language (human and machine readable)
-
3
Transform state information and formal ruleset into geometry of the compliant area.
-
4
Check if the nominal path is in the compliant and safe area.
-
5
Emergency path calculation
Kontrol KoSim is a simulation environment
for validation of autonomous systems in the
development phase.
With Kontrol KoSim our customers can simulate scenarios without a large investment of time or money.
- Enable formal verification during development
- Reduce costs with early validation in development phase
- Fully customized rule library
– Select and configure dynamic model
– Set parameters
– Build legal tree
– Set up parameters for the rules
– KoPilot: Recorded data
– KoSim: Simulation data
– Boolean
– Probabilistic
– KoPilot: Rule violations
– KoSim: Data before rule violation
The international research and development centre Virtual Vehicle and Kontrol have further advanced a possible certification of ADAS/AD at the proving ground of ÖAMTC Fahrtechnik Zentrum Lang/Lebring.
Kontrol Smart Lib technology logically interprets and transforms applicable regulations (e.g. traffic act) into computable code.
The Kontrol Check algorithm compares if a planned vehicle trajectory violates any of the boundary conditions computed by the Kontrol monitor.
Kontrol Monitor predicts the allowed reachable area for the ego vehicle for the next iteration step. The allowed reachable area is based on the current vehicle state, a physical model of the car, and the boundary conditions, derived from the Kontrol Smart Lib.
Kontrol Safe is an operational fail safe mode for autonomous vehicles. If the Kontrol
Check algorithm detects a violation of the boundary conditions (e.g. rule violation), an alternative trajectory or emergency maneuver can be enforced.
With these four modules the Kontrol KoPilot enables the autonomous vehicle to calculate the reachable area to apply all relevant rules and to verify the path.