From Footfall to Foresight: Revolutionizing Retail Analytics with Real-Time Edge AI

White Paper Overview

In a world where physical presence is as valuable as digital engagement, understanding how people move through spaces is critical. Traditional methods of gathering this data—manual counting or cloud-based video analytics—are either inaccurate or raise privacy and cost concerns. This white paper introduces a real-time, edge-based people counting solution developed by SECO and powered by NXP’s i.MX 95 application processors. It showcases how AI at the edge can deliver accurate, GDPR-compliant insights without relying on cloud infrastructure.

The solution runs entirely on-device, using SECO’s SOM-SMARC-MX95 and NXP’s eIQ Neutron NPU to process video locally. It ensures privacy, reduces latency, and simplifies deployment through containerized architecture and Clea OS. You’ll learn why traditional analytics fall short and how edge AI solves key issues, explore the hardware and software powering SECO’s solution, and see how it was deployed at Embedded World 2025 with applications in retail, smart buildings, transportation hubs, and more. The white paper also explains how Clea enables fleet-wide management, secure updates, and predictive analytics.

Key takeaways:

• Privacy-first design: No images or biometric data stored; only anonymized metadata is used.
• Real-time insights: Live dashboards visualize occupancy and trends.
• Scalable architecture: Docker-based microservices simplify updates and deployment.
• Cross-industry value: From optimizing store layouts to managing crowd flow in stadiums, the solution adapts to diverse environments.

Published October 2025 by SECO and NXP Semiconductor