Le plafond de 25 km/h des EDPM à l'épreuve des accidents recensés
Abstract
The 25 km/h construction speed cap for powered Personal Electric Vehicles (PEVs) was set in France in 2019 "by analogy with the pedal-assist threshold of electrically power assisted cycles (EPACs)", not on the basis of crash data specific to the mode. This article tests this threshold against bodily injury crashes recorded in France from 2019 to 2024, by matching the full geocoded BAAC police file to OpenStreetMap functional road classes (coverage above 99%). Two questions are asked: do PEVs show, at comparable context, an excess vulnerability to road surface features that would justify a single speed cap, and does a single cap describe the actual insertion of these devices in traffic? The analysis covers 10 298 PEV crashes (including 343 outside built-up areas), compared with bicycles through standardisation on speed limit, site and surface. The broad composite of vehicle–surface interactions is nearly identical across modes (9.97% vs. 9.67%): the data do not confirm a general excess vulnerability. However, non-collision crashes (10.55% vs. 7.97%) and kerb strikes are more frequent among PEVs, and the share of crashes on adverse surface outside built-up areas (18.95% vs. 10.36%) is the only statistically separated gap in the design. Outside built-up areas, 323 crashes are recorded off cycleways under a default prohibition regime, rising from 9 in 2019 to 120 in 2024. We conclude that the 25 km/h cap is not a standalone safety criterion: neither its removal nor its maintenance follows from the observations; it should be replaced by a joint assessment of device, road and traffic, within a regulated opening of low-traffic rural roads.
Related
BibTeX
@misc{lyan2026plafond,
title = {{Le plafond de 25 km/h des EDPM à l'épreuve des accidents recensés}},
author = {Gauthier Le Bartz Lyan},
howpublished = {HAL},
year = {2026},
month = oct,
url = {https://inria.hal.science/hal-05784630v1},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Le plafond de 25 km/h des EDPM à l'épreuve des accidents recensés
AU - Gauthier Le Bartz Lyan
PY - 2026
DA - 2026/10/07
PB - HAL
UR - https://inria.hal.science/hal-05784630v1
AB - The 25 km/h construction speed cap for powered Personal Electric Vehicles (PEVs) was set in France in 2019 "by analogy with the pedal-assist threshold of electrically power assisted cycles (EPACs)", not on the basis of crash data specific to the mode. This article tests this threshold against bodily injury crashes recorded in France from 2019 to 2024, by matching the full geocoded BAAC police file to OpenStreetMap functional road classes (coverage above 99%). Two questions are asked: do PEVs show, at comparable context, an excess vulnerability to road surface features that would justify a single speed cap, and does a single cap describe the actual insertion of these devices in traffic? The analysis covers 10 298 PEV crashes (including 343 outside built-up areas), compared with bicycles through standardisation on speed limit, site and surface. The broad composite of vehicle–surface interactions is nearly identical across modes (9.97% vs. 9.67%): the data do not confirm a general excess vulnerability. However, non-collision crashes (10.55% vs. 7.97%) and kerb strikes are more frequent among PEVs, and the share of crashes on adverse surface outside built-up areas (18.95% vs. 10.36%) is the only statistically separated gap in the design. Outside built-up areas, 323 crashes are recorded off cycleways under a default prohibition regime, rising from 9 in 2019 to 120 in 2024. We conclude that the 25 km/h cap is not a standalone safety criterion: neither its removal nor its maintenance follows from the observations; it should be replaced by a joint assessment of device, road and traffic, within a regulated opening of low-traffic rural roads.
ER -