Robotic Weld Cells Need Instant Thermal Insights
HeatCore™ Robotic integrates cameras, AI, and control interfaces so every weld seam is verified in-line—without slowing the robot down.
See it in action
Thermal defects stay hidden until post-process inspection
Robots weld at speed, but teams usually inspect later. Porosity, burn-through, or heat drift only appear after cutting, machining, or—worse—at the customer site.
Typical pain points
- No thermal visibility during the weld cycle
- Defects are found after value has already been added
- Robots stop for manual verification and parameter tweaks
The business impact
- Scrap and rework inflate costs and kill takt time
- Audits take too long because evidence isn’t real-time
- Maintenance teams react late to torch misalignment or gas issues
Inline thermography + AI tuned for robots
The system synchronizes thermal cameras, edge AI, and operator interfaces so every welding cell can self-verify each bead in real time.
Synchronized thermal cameras
Mount directly on robot wrists or fixtures. Capture 60 FPS thermal video with filters that block arc glare.
- IP67-rated housings with air purge options
- Monitors temperature gradients and cooling rates
- Robot controller integration verified during project engineering
AI decision engine
HeatCore AI™ classifies porosity, lack of fusion, and overheating on the fly, flagging suspect welds before they leave the cell.
- 99%+ detection accuracy when paired with multi-modal arc and thermal sensing
- Sub-100 ms feedback loop
- Adaptive thresholds per joint and recipe
Panel PC + dashboards
Operators see live thermal overlays, alerts, and parameter drift. Supervisors review dashboards per cell, line, or plant.
- Edge UI with glove-friendly controls
- Auto exports to QMS Copilot™ and MES
- Historical analytics for maintenance and process engineering
See the line in action
Related products
See the AI engine that powers the cell, the entry-level tier, and the full system overview.
HeatCore tiers
System overview
Frequently asked questions
What is HeatCore Robotic?
HeatCore Robotic is a turnkey weld monitoring cell that integrates thermal cameras, edge AI, and operator interfaces so a robotic welding station can verify each bead inline, in real time, instead of relying on post-process inspection.
Which robot brands does HeatCore Robotic work with?
The thermal cameras and control interfaces are designed to work with common industrial robot controllers through their digital I/O. Final integration depends on your cell layout, controller, and I/O.
Where are the thermal cameras mounted, and can they handle arc glare?
Cameras mount directly on robot wrists or fixtures and capture thermal video with filters that block arc glare. Housings are IP67-rated with optional air purge for demanding shop-floor environments.
What weld defects can the cell flag inline?
The HeatCore AI engine classifies thermal patterns correlated with porosity, lack of fusion, and overheating, flagging suspect welds before they leave the cell. Coverage depends on material, joint geometry, and the acceptance criteria configured for your process.
Does inline monitoring slow the robot down?
No. The system synchronizes cameras, edge AI, and operator interfaces so verification runs during the weld cycle without stopping the robot, with a sub-100 ms feedback loop and adaptive thresholds per joint and recipe.
Can HeatCore Robotic integrate with our MES or ERP?
Yes. Results auto-export to Therness QMS Copilot and MES, and supervisors can review dashboards per cell, line, or plant. Historical analytics support maintenance and process-engineering follow-up.
What do operators and supervisors see during production?
Operators work from an edge panel PC with live thermal overlays, alerts, and parameter-drift indicators using glove-friendly controls, while supervisors review aggregated dashboards and historical trends.
Scope HeatCore Robotic for your line
Send the joint, the process and the cell constraints. A Therness engineer replies within one business day with a deployment path, not a sales deck.
