In 2015, Christophe Bastien collaborated with West Midlands Police (WMP) to investigate the complex relationship between vehicle damage, pedestrian injury, and post-mortem findings in hit-and-run cases. By collecting and analysing pedestrian collision data from the WMP Central Motorway Police Group (CMPG), he demonstrated that trigonometric methods, combined with a bird’s-eye view representation of the vehicle, could establish a link between vehicle impact speed, vehicle damage, the pedestrian’s point of contact, and leg length.
He advanced this work further by mathematically demonstrating that pedestrian crossing speed could also be derived from the same information. This extraction of crossing speed represented a significant forensic innovation that had often been overlooked. Since impact speed is directly correlated with injury severity, this finding suggested that key elements of hit-and-run cases could potentially be reconstructed using only vehicle damage and pedestrian medical records.
In 2018, Christophe published The Pedestrian Crossing Speed Calculator (PCSC), presenting this methodology. However, despite its mathematical accuracy, the approach proved difficult to apply in practical forensic contexts. The PCSC relied on interpreting the pedestrian’s gait at the moment of impact based on permanent bonnet deformation, which introduced significant uncertainty. As a result, the method was not considered sufficiently reliable for routine forensic use. The collaboration with WMP eventually ended due to staff reductions resulting from government cutbacks.
Between 2019 and 2021, Christophe focused on pedestrian brain injury severity modelling. In 2021, he and his team were awarded the Prince Michael International Award for Road Safety for implementing a new brain injury criterion based on theoretical work by Professor Clive Neal-Sturgess. This achievement effectively addressed the post-mortem injury component of the earlier hit-and-run challenge and reignited Christophe’s interest in solving the remaining problem of impact speed determination.
To overcome the limitations of PCSC, Christophe developed a new approach that no longer depended on pedestrian leg length alone. While maintaining the principle of extracting impact speed from vehicle damage, he adopted human computational models capable of reproducing realistic three-dimensional kinematics during vehicle collisions. This marked a shift in paradigm—from simplified geometric assumptions to full human anthropometry (including height and weight) combined with simplified three-dimensional vehicle models.
With the support of his PhD student, Dr Vadhiraj Shrinivas, Christophe generated thousands of simulated training cases. These datasets explored pedestrian kinematics as a function of multiple variables, including vehicle profile, damage characteristics, pedestrian height and weight, gait, crossing angle, crossing speed, and head contact location. The resulting data were used to train an artificial intelligence system capable of making forensic collision predictions: the Pedestrian Collision Evaluator (PACE-AI).