Driving zero-defect manufacturing through edge-native, label-free anomaly detection — on every line, at every site.

Our vision is to create a unified Machinetra ecosystem where QPath seamlessly connects cameras, robots, and operators — transforming inspections into a continuous, self-improving quality journey.

• Edge-native anomaly detection (vision & audio)
• QPath: integrated web platform + edge app
• Modular hardware kits for fixed & mobile deployments

Self‑POC

Upload data, train unsupervised models, validate results, and move to production in hours.

• Drag‑and‑drop datasets
• Auto‑tuned parameters
• Versioned experiments

Edge Fleet Management

Register devices, push models, configure pipelines, and monitor health on a live map.

• Device provisioning
• Model rollouts
• Live telemetry

Reporting & Analytics

Turn inspections into insight: exportable dashboards, KPIs, and audit trails.

• Per‑line KPIs
• False‑positive tracking
• CSV/PNG/PDF export

Key capabilities

• Domain‑independent unsupervised detection (no labeled defects)
• <50 ms per‑frame latency on Jetson‑class devices
• On‑device feedback loop — no full retraining cycles
• Configurable pipelines (pre/post‑processing, thresholds, alarms)

Integrations

• GigE/USB/industrial cameras, PTZ control
• Digital I/O, OPC‑UA/MQTT for line signaling
• REST/ROS2 API for robots and MES/SCADA
• Role‑based access & secure device updates

Fixed Stations

Compact edge boxes with industrial cameras for in‑line inspection.

• GigE/PoE
• IP‑rated enclosures
• Vibration isolation

Mobile (AGV/UGV)

Autonomous patrols for multi‑point or preventive inspections.

• Navigation + PTZ
• Hot‑swap batteries
• Docking & auto‑upload

PTZ / GigE Kits

Drop‑in kits to upgrade existing cameras with edge AI.

• Jetson/ARM/Coral
• I/O & field‑bus
• Remote provisioning