草榴社区

Introduction to PEGASUS and Autonomous Driving

The much-discussed research project PEGASUS from Germany laid out methods for the testing and development of automated driving systems. Only a few years after its inception it has become clear that the PEGASUS method is here to stay. Testing autonomous driving systems with the PEGASUS method can be readily employed with TPT which is fully equipped to join you on the (automated) test ride.

The PEGASUS Method: A Scenario-Based Approach

Research revolving around autonomous driving essentially deals with the question whether a machine can drive better and more safely than a human. In order to find an answer, extensive testing has to be done and a myriad of aspects need to be taken into account. But how do you test? Driving millions of kilometers (or miles) did not yield satisfactory results and entailed considerable time and cost. Instead, the PEGASUS method is based on a scenario-based approach in order to define the “new general state of the art for test methods and test case selection” (, p. 4)

What’s more, the PEGASUS method suggests six layers that describe the scenarios on the road systematically:  

1. Road-level (geometry, …)
2. Traffic infrastructure (traffic signs, …)
3. Temporary manipulation of layer 1. and 2. (road construction, …)
4. Objects (moving relative to vehicle under test: pedestrians, …)
5. Environment (weather, …)
6. Digital information (e.g. V2X information)

Focusing on layers makes scenario-based testing viable in a structured and straightforward fashion. This way you can identify what circumstances you need to create in order to test automated driving systems.