Logistics & TransportationCase study · Multiple sites

Computer vision object detection across a logistics network

Duration: 6 months · Team size: 5–8 specialists

Get in touch

The challenge

A logistics organisation needed real-time monitoring of people and vehicle movements using its existing camera network, for operational optimisation and safety. Cameras were passive recorders only, footage review was manual, privacy compliance required face blurring, and connectivity varied: some sites had cloud access while others were remote and needed edge processing.

What we did

Developed a framework combining computer-vision detection, automation, cloud storage and dashboarding, supporting both cloud and edge deployment modes across sites with different connectivity.

The outcome

Enabled near-real-time monitoring of people and vehicle events with structured analytics, delivered as a reusable framework offering operational transparency and actionable metrics.

Turned a passive camera network into near-real-time, privacy-compliant operational analytics across cloud and edge sites.

Databricks products used
Model ServingMLflowDelta LakeDatabricks Workflows
Capabilities applied
Computer vision object detectionEdge & cloud deploymentPrivacy-compliant processingOperational analytics
Technical depth

The framework ran the same detection models at the edge or in the cloud depending on site connectivity, with face-blurring applied before storage and structured event data surfaced through dashboards.