Inovative TechnologiesInovative Technologies

Manufacturing

Vision-based QC — fewer false rejects, faster lines.

We implemented a computer-vision quality control system across two production lines. Models run at the edge, pipeline data fuels continuous re-training, and supervisors get near-real-time OEE visibility.

Vision QC cover
−31%
False rejects
−19%
Cycle time
+12%
Throughput
<120 ms
Median inference latency

Overview

Operators were over-rejecting good units and re-inspecting borderline cases. We deployed a vision system that flags defects with explainable overlays and collects image samples to improve future models. The result: fewer unnecessary stops and smoother flow.

  • Edge-first inference with GPU acceleration
  • Operator UI with overlays and quick accept/override
  • Automated data curation → re-training → safe rollout
  • OEE/ SPC dashboards for supervisors & quality
  • Full audit trail and model version traceability
Vision QC workflow

From camera to decision to monitoring & improvement loop.

Approach

Pilot & baselines

Instrument line cameras, collect labeled sets, measure baseline false reject rate.

Modeling & validation

Augment, train, calibrate thresholds; human-in-the-loop QA.

Edge deployment

Optimize to ONNX/TensorRT; health checks, fallbacks, offline buffers.

Operator UX

Head-up defect overlays, quick actions, feedback capture.

MLOps & monitoring

Drift alerts, shadow tests, canary rollouts on shifts.

Scale out

Golden image, IaC and device fleet management across sites.

Tech stack

  • Edge inference (ONNX/TensorRT), GPU/Jetson deployment
  • Training: PyTorch, Albumentations, Weights & Biases
  • Data pipeline: Kafka, S3/Lake, Delta/Parquet
  • MLOps: MLflow, feature store, model registry
  • Monitoring: Prometheus, Grafana, OpenTelemetry
  • OEE dashboards, SPC alerts, traceable audit logs
Line performance (pilot → scale)
Performance trend

Steady reduction in false rejects after model iterations; throughput lift with latency kept under 120 ms.

“Operators trust the overlays, supervisors trust the dashboards. Quality meetings are now about trends, not screenshots.”

— Plant Quality Manager

Ready to reduce rejects?

We can pilot on one line within weeks, then scale across sites with confidence.