Roermond 'Sensing Project' predictive policing

Occurred: October 2020

A predictive policing system in Roermond, the Netherlands, has been castigated for being discriminatory, an abuse of privacy, ineffective, opaque, and unaccountable.

Operating between January 2019 and October 2020, the 'Sensing Project' used cameras and sensors to collect data on vehicles driving in and around the city, supposedly to reduce shoplifting and pickpocketing. 

The data was then assessed by an algorithm that calculated the probability that the driver and passengers intend to engage in forms of 'mobile banditry' such as pickpocketing and shoplifting, and directs police towards the people and places it deems 'high risk', according to Amnesty. 

Marketed as a neutral system that uses objective crime data, Amnesty discovered that it is designed to identify people of Eastern European origin, notably those of Roma ethnicity, and exclude local nationals. 

Amnesty also found that the system created many false positives, that the police were unable to demonstrate its effectiveness, and that no one in Roermond had consented to its use.

Operator: Roermond Municipal Council

Country: Netherlands

Sector: Govt - police

Purpose: Reduce crime; Profile ethnicity

Technology: Machine learning; Pattern recognition
Issue: Accuracy/reliability; Bias/discrimination - race, ethnicity; Effectiveness/value; Surveillance; Privacy

Transparency: Governance; Marketing

Research, advocacy 🧮

Page info
Type: System
Published: September 2023
Last updated: November 2023