In order to improve the preciseness in pedestrian-vehicle accident reconstructions, a new biofidelic dummy had been developed. The objective of this work was to biomechanically validate the biofidelity of this new kind of anthropomorphic testing device. Therefore, nine crash tests have been conducted with the biofidelic dummy and the results were compared with four crash tests earlier performed with the Žilina dummy, an anthropomorphic testing device widely used in accident research due to its low cost and robustness, cadaver tests obtained from published research papers and 21 real-world pedestrian accidents.

The trajectories of both anthropomorphic testing devices were computed and compared with those of cadaver tests, highlighting that the biofidelic dummy performs much more human-like than the Žilina dummy.

Damages to the vehicle’s front caused by both anthropomorphic testing devices and real pedestrians were compared with each other, as realistic damages are very important for reconstruction purposes. It can be shown that the biofidelic dummy causes damages similar to those a pedestrian would cause in an accident of similar severity, whereas the Žilina dummy causes damages which are far too severe, which may mislead the expert witness to assume an impact velocity which is too slow.

The C-ratio, defined as closing speed over collision speed, was computed for both anthropomorphic testing devices, showing that the two deliver similar results. Computing the dynamic, time-dependent C-ratio, however, highlights differences in the kinematics and dynamics of the two anthropomorphic testing devices.

The throw distances of both anthropomorphic testing devices were compared with throw distance charts developed in-house by DEKRA based on experiments with the Žilina dummy and well-documented real-world accidents. The results show that the biofidelic dummy’s behaviour is good.

Finally, the damages to the biofidelic dummy were analysed and transcribed to injuries of a human being. These “biofidelic dummy injuries” were compared with injuries of real pedestrians, focusing on five injuries, which can be used for reconstruction purposes. In general, the “injuries” of the biofdelic dummy correspond well with those of the pedestrian.