A French Tech Company Is Showing Europe How AI and GDPR Can Work Hand in Hand

As the European Union continues to refine its AI Act and data protection standards remain among the strictest in the world, a quiet but telling example of responsible AI in practice is emerging from the public transport sector.
Acorel, a French technology company based in Saint-Péray, has developed a computer vision system for public transit networks that delivers real-time operational intelligence without capturing, storing, or processing any personal data. No facial recognition. No video recordings. No individual tracking of any kind. Just anonymised passenger flow data, processed in real time and fed directly to transit operators.
The timing is significant. Brussels finds itself at the centre of a global debate about where to draw the line between the promise of AI-driven public services and the fundamental right to privacy. Critics of smart city technology frequently point to surveillance creep — the gradual expansion of monitoring infrastructure under the banner of operational necessity. Acorel’s approach offers a pointed counter-argument: that it is entirely possible to build effective AI systems without that trade-off.
The technology works through overhead counting sensors installed inside buses, trams, and trains. These sensors track boardings, alightings, occupancy levels, and crowd dynamics as they happen, giving operators a continuous, accurate picture of how their network is performing. Fleet managers can identify overcrowded routes before congestion builds, reallocate vehicles in real time, and base longer-term planning on reliable ground-level data rather than modelled projections — all without a single surveillance camera on the network.
The architecture is built around GDPR compliance from the ground up, treating data minimisation not as a legal checkbox but as a foundational design principle. For European public authorities navigating procurement decisions in an era of heightened regulatory scrutiny, that distinction matters.
“Computer vision can be a powerful optimization tool when it is designed responsibly,” said Dimitri Rudenko, Business Development and Project Director at Acorel. “Efficiency and data protection are not contradictory — they are complementary.”
The system is designed to operate accurately in variable lighting and high-density conditions — precisely the peak-hour environments where occupancy data is most operationally valuable. It is compatible with buses, trams, trains, and a range of other public transport formats.
As European institutions debate how to govern AI deployment in public spaces, examples like this one carry a broader significance. They demonstrate that the choice between effective public services and citizen privacy is not binary — and that European technology companies are already building the alternatives.
Further information is available at acorel.com or by contacting mobility@acorel.com.