Edge gateways
Inference next to the camera or sensor, for sub-second decisions at the station.
On-premiseConvert manual, paper-based or spreadsheet-driven shop-floor processes into AI-assisted operations, with human-in-the-loop governance, audit trails and measurable throughput gains.
Generates refrigerant circuit (CKT) routing layouts for fin-and-tube heat exchangers from the coil grid, target circuits, parity, flow direction and dummy holes. It applies the design team's bend, drainage and coverage rules, balances circuit sizes across the face, and exports an SVG/PDF reference with every run saved for audit and reuse.
Agentic re-planning that proposes revised schedules when an order changes. A plain-language interface lets planners request a build plan or dispatch report, checked against the BOM and live stock before anything is committed.

Coil circuit routing that used to be drawn by hand is now generated by rules, checked, and exported as SVG/PDF, with every run saved for audit and reuse. Delivered to a heat-exchanger OEM and running fully offline on their own machines.
Cameras and sensors feed edge gateways or line-side compute. Models, the copilot and the audit log live on your plant server or in your private cloud. Alerts and work orders flow back into the systems you already run.
Public AI services: nothing crosses this line. No server-side logging of proprietary parameters, no training on your production data, no third-party broker in the pipeline.
AI that runs within your firewall, with your total control. On-premise or private cloud is your call, and we engineer for either.
Inference next to the camera or sensor, for sub-second decisions at the station.
On-premiseOne compact node per line, running inspection, counting and monitoring together.
On-premiseCentral models, dashboards and the copilot, inside your own network and under your own IT policies.
On-premiseYour own cloud tenancy, inside your security perimeter, when that suits your IT strategy better.
Private cloudThe AI proposes; planners and engineers approve before anything is committed.
Inputs, outputs and decisions are stored for traceability and reuse.
Your team's design and business rules are enforced on every output, with deviations explained.