Task

The task was to count individuals ascending and descending stairs in a subway station while fully adhering to all applicable data protection regulations. The objective was to generate reliable pedestrian flow data to sustainably improve safety, planning, and resource management in public transit facilities.

Challenge

Accurately measuring pedestrian flows in public transit stations is critical for identifying and mitigating capacity bottlenecks, safety hazards, and emergency situations at an early stage. However, stringent privacy regulations pose challenges, requiring that personal image data neither be stored nor transmitted. Local data processing and secure aggregation of anonymous counting data are thus imperative.

Solution

SALTIR implemented a privacy-compliant, AI-based computer vision solution directly at a subway staircase for pedestrian counting. Image processing occurs exclusively on the local device. Only anonymized counting data is securely transmitted to central servers. Image data never leaves the device and is not stored, thereby ensuring the highest standards of data privacy.

Results

The practically implemented solution enables precise, privacy-secure analysis of pedestrian flows in real-time. Municipalities and transit authorities benefit from improved planning, increased safety, and optimized resource management. The solution is easily scalable, making it highly suitable for widespread deployment in urban infrastructure.