Scenarios
The derivation of relevant scenarios from accident statistics directly is a challenge in itself. No reliable accident databases are available that are capable of delivering a comprehensive analysis of accident circumstances for the whole of Europe. While some European projects such as TRACE have been working on ideas for the harmonisation of accident statistics, waiting for them being available is not acceptable. Thus the partners have defined relevant scenarios based on information that is available today. This includes standards for testing of certain systems, results from other projects and the expertise of the involved institutions.
Cluster 1: Longitudinal Assistance
For System Cluster 1, three different scenarios have been chosen. They represent a straight road, a curved road and a target, which is transversally moving in the way of the subject vehicle.
Regarding the straight road, the objective of the chosen scenario is to validate that the subject vehicle can detect and handle (warn, support, and/or intervene) a target vehicle in the same lane:
Straight Road Scenario
The same objective applies for the scenario, however for a curved road:
Curved Road Scenario
The objective of the third scenario is to validate that the subject vehicle can detect and handle (warn, support, and/or intervene) a target (e.g., other vehicle, pedestrian, …) which moves lateral to the subject vehicle:
Transversally Moving Target Scenario
Cluster 2: Lateral Assistance
The System Cluster 2 is addressing systems which are providing lateral assistance. For straight as well as curved roads, a differentiation is made regarding lane and road departure. Accordingly, four different scenarios are considered.
The first scenario is meant to validate the subject vehicle capability to avoid involuntary (left/right) lane departure driving on a straight road:
Lane Departure on a Straight Road Scenario
As a form of extension of the first scenario, the second is meant to validate the subject vehicle capability to avoid involuntary road departure driving on a straight road:
Road Departure on a Straight Road Scenario
Comparable to the first two, the second and third scenario of Cluster 2 regards lane or road departure while the subject vehicle is driving in a curve. Again, the capability to avoid the involuntary lane or road departure is the objective here:
Lane or Road Departure in a Curved Road Scenario
A modification to the aforementioned is given by scenario five and six, namely to validate the subject vehicle capability to avoid involuntary lane departure driving on a straight road just before entering an upcoming curve:
Lane or Road Departure on a Straight Road Just Before a Curve Scenario
While these scenarios do not consider interaction with a second (called target) vehicle, the seventh scenario does so. It addresses lane change collisions which are well-known in multi-lane traffic both at low and high speeds:
Lane Change Collision Avoidance on a Straight Road Scenario
Cluster 3: Yaw/stability Assistance
Yaw and stability assistance is given by systems which have been collected under System Cluster 3. Here, some manoeuvres are already established in testing. One example is braking on μ-split, i.e. surfaces with different friction coefficients:
Emergency Breaking on µ-Split Scenario
The capability of the vehicle to avoid loss of control in a sudden obstacle avoidance manoeuvre is regarded with the second scenario in Cluster 3:
Driver Collision Avoidance Scenario
Finally, critical situations linked to curved roads are represented by the third and fourth scenario of Cluster 3:
Fast Driving into a Curve Scenario
Roll Stability Scenario
All scenarios do not only consider passenger cars but generally also apply for trucks and busses. However, it has not been decided yet to what extent the project can regard the special requirements by commercial vehicles concerning active safety test methods.
Next Steps
Having defined the scenarios, the development of the methods themselves has been started. The main focus will be on physical testing with a certain support from simulation where this seems appropriate.