byASB Ankiit Singh Book a scoping week
POC ✓2019–21

State-scale ANPR and face recognition

A state police force (agency not named)
Large-scale deployment architecture under real constraints. Successful proof-of-concept conducted.
Scale1,400 cameras
InferenceEdge, per camera · Jetson-class
ReferenceLive six-camera deployment
DeliveredArchitecture + 3-year cost model
VehicleDelivered under Utenx
Try the TCO model ↗
The constraint

A 1,400-camera video platform designed the obvious way — stream everything to a central GPU farm, archive all of it — hits a bandwidth and cost wall long before it hits a technical one. The architecture has to be economically survivable for three years, not just demonstrable once.

What I built

Architected the technical approach for a state-scale law-enforcement video platform: automatic number-plate recognition and face recognition across 1,400 cameras, with edge inference per camera rather than centralised GPU processing, event-triggered retention rather than full archival, and integration into national vehicle and identity databases.

Delivered the full cost and architecture model — three-year TCO, per-camera economics, and the build-versus-buy analysis — alongside a live six-camera reference deployment.

Under the surface
inferenceedge, per camera · Jetson-class · no central GPU wall
retentionevent-triggered · not full archival
economics3-yr TCO · per-camera all-in · benchmarked vs Safe City
reference6-camera live POC · ANPR + FRS
compliancedata-protection framing · certification path
exposure: state scale · 24×7 · law-enforcement critical
Outcome
  • Successful six-camera proof-of-concept.
  • Full three-year TCO and per-camera economics delivered.
  • Build-versus-buy analysis benchmarked against Safe City deployments.
Jetson-class edge inferenceANPRFace recognitionEvent-triggered retentionNational DB integration