The vision AI platform built for the line, not the lab.

Teach from a handful of good parts. Detect anomalies nobody labelled. Trace each defect class to its cause. Deploy on the edge, inside your plant network.

Get started
LIVE
Simulated feed
inference
pass 0anomaly 0rate 0.0%
edge · on-prem · 0 frames uploaded
Inspection log
  • Waiting for parts…
Root-cause signal

No correlated pattern yet. Watching Line 3 for process drift.

A dozen good parts is a training set.

Onqyra ships with a vision foundation model pretrained on industrial imagery — surfaces, packaging, webs, castings, prints. Teaching a new part is fine-tuning that prior on a handful of your own good samples, captured on your line with your lighting. The result is a versioned model of “normal” for that SKU, ready to deploy in minutes.

  • Self-supervised: no defect labels, ever
  • Handles natural variation in food, textiles and raw materials
  • Versioned, staged, reversible per camera and per line
teach · blister-pack-v14simulated
good_01.png
good_02.png
good_03.png
good_04.png
good_05.png
good_06.png
good_07.png
good_08.png
good_09.png
good_10.png
good_11.png
good_12.png
Model ready
12 samples · self-supervised · no labels required
Deploy to Line 3

If it deviates from normal, it lights up.

Every frame is scored against the learned representation. Deviations are localised to the pixel, scored, and clustered. A defect type nobody has seen before still registers on its first occurrence — then Onqyra groups recurrences so QA can name the class in one click and track it from then on.

  • Pixel-level anomaly maps with confidence scores
  • Auto-clustered defect classes, human-in-the-loop naming
  • Thresholds per SKU, per camera, with drift monitoring
detect · line 3 · cam 02simulated
anomaly map · pixel-levelscore 0.97
  • Missing tablet
    named by QA
    0.97
  • Foil scuff
    named by QA
    0.91
  • Unclassified cluster #7
    never seen before
    0.88
  • Seal void
    named by QA
    0.84

Defects, correlated to the process that made them.

Each anomaly carries the state of the line: shift, tool, batch, supplier lot, upstream sensor values. Onqyra runs continuous correlation across defect classes and process events, surfaces lifts as they emerge, and publishes signals your PLC and MES can act on.

  • Correlation across line, shift, tool change, lot, sensor state
  • Signals over OPC-UA, MQTT, Modbus TCP, REST
  • Scrap and warranty impact reported per defect class
trace · foil scuff · last 72 hsimulated
Where “Foil scuff” shows up
lift vs. plant baseline
  • Line 3 · after 06:40 tool change3.4×
  • Supplier lot B-118 foil2.1×
  • Night shift · sealer temp > 182 °C1.6×
  • Line 1 · baseline1.0×
Signal pushed · OPC-UA

Foil scuff correlates with Line 3 after tool change. Recommend sealer roller inspection before next changeover.

Under the hood.

A foundation model, an anomaly engine and a fleet manager — packaged for people who run lines, not GPUs.

Industrial vision foundation model
Pretrained on surfaces, packaging, webs and castings so few-shot adaptation converges fast and robustly.
Anomaly scoring engine
Patch-level representations compared against the learned normal manifold; localisation without segmentation labels.
Active-learning loop
Operator confirmations feed back into per-SKU thresholds and defect-class definitions without retraining from scratch.
Model registry & fleet manager
Every model versioned with its samples and metrics. Stage on one camera, promote to a line, roll back in a click.
Edge runtime
Optimised inference on the Onqyra edge appliance, keeping pace with the line without a cloud round-trip.
Process integration
Native OPC-UA and MQTT clients, Modbus TCP, REST and webhooks. Historian and MES connectors on request.
Audit-grade logging
Verdicts, frame hashes, model versions and signals retained and exportable for QA and regulatory review.
Roles for the plant
Operators confirm, QA names classes, engineers wire signals, managers watch scrap fall. One workspace.

Runs on the line. Stays in the plant.

Inference happens on an edge appliance next to the camera, not in someone else's cloud. Your frames, your models, your network — with the connectivity a real plant floor demands.

deployment topologyplant network · VLAN 40
Cam 01
GigE · line 3
Cam 02
GigE · line 3
Cam 03
USB3 · line 4
Onqyra Edge
on-device inference
PLC / SCADA
verdict · OPC-UA
MES / historian
root cause · MQTT
0 frames leave the plant
Any industrial camera
GigE Vision, USB3 Vision, line-scan or area-scan. Keep the optics you already own.
Edge-native inference
Models run on the Onqyra edge appliance beside the line, at line speed, with no cloud round-trip.
Frames never leave the plant
Images stay on-prem by default. Only anonymised metrics and model deltas sync — or nothing at all, air-gapped.
Talks to what you run
OPC-UA, MQTT, Modbus TCP, PROFINET via gateway, and REST for MES, SCADA and historians.
Fleet-managed models
Teach on one line, roll out to twelve. Every model versioned, staged and reversible.
Built for audit
Every verdict, frame hash and root-cause signal is logged and exportable for QA and regulatory review.

Put an inspector on every line that never blinks.

Pilots start with one camera and a dozen good parts. Most plants see scrap move within the first weeks.

Get started See pricing