Abstract
Objective
From 2018, AEB systems dedicated to avoid or mitigate car-to-cyclist collisions will be considered in the safety assessment by Euro NCAP. To test such systems, appropriate equipment is being developed in the project CATS “Cyclist-AEB Testing System.” Moreover, the project dealt with setting up a proposal for the most relevant test scenarios. The objective of the project was to provide proof of the relevance of the proposed test scenarios and of the feasibility of practical implementation of the scenarios and test setup to Euro NCAP.
Methodology
Accidentology was used to determine the most common car-to-cyclist accident scenarios in the EU. The test setup has been developed to deal with the 3 most relevant scenarios. Apart from the scenarios, the typical AEB systems and the applied sensors should recognize the dummy as a real cyclist on a bike. In this way, requirements regarding dummy visual and radar characteristics have been set up. Moreover, requirements to the design of the testing system and dummy result from practical testing constraints: during testing, possible collisions between car and dummy should be mild without compromising the testing. A dummy and dummy propulsion system was constructed to meet these requirements. Accident data and data from observation studies were used to determine the parameter ranges in the proposed test matrix for the selected test scenarios.
Results
Together with car manufacturers and suppliers, a test matrix has been proposed and a testing setup including cyclist dummy has been built. In several test series, the test setup and the protocol were verified. Verification considered not only the capability of the protocol to distinguish between well-performing and less-performing AEB systems, but also the reproducibility and repeatability of the tests. Also the practical use of the system and the dummy were tested. This resulted in a protocol and test matrix for 3 scenarios: a cyclist crossing the path of the car from the near side, a cyclist crossing the path of the car from the far side, and a longitudinal scenario in which the car and cyclist are driving in the same direction, and the car drives into the cyclist from the rear. These scenarios cover 63 to 78% of the car-to-cyclist accidents with respect to the number of seriously injured and fatalities respectively.
Limitations of the study
The availability of accident data regarding cyclist accidents and naturalistic bicyclist data is limited. Detailed accident databases, required for the CATS analysis, were only available for 6 out of the 28 EU countries. Nevertheless, most countries in which the popularity of cycling is high or strongly increasing (France, Germany, Italy, the Netherlands, Sweden and the UK), were included in the study. The cyclist target resulting from CATS has rotating wheels, but no articulation regarding legs, head or torso.
What is new
The CATS project was unique in the fact that the developments were performed in a consortium of 10 car manufacturers and 7 automotive suppliers, with BASt as review partner. This has very much stimulated the harmonization and acceptance of the protocol, dummy and test setup. The progress and intermediate results including the used methodology have been shared on a regular basis during the project with stakeholders in Europe, Japan and the USA.
Conclusions
Euro NCAP already indicated to consider the results of the CATS project as the main input to draft the test protocol, including scenarios and target, for Cyclist-AEB systems in 2018.