NFO plugs into the PLCs you already run — Modbus or OPC UA — reasons over every signal in real time, and hands your operators maintenance calls with the evidence attached. Entirely on your hardware.
Everything above is simulated — a working slice of the real console, running on synthetic data in your browser. In production, NFO reads your own PLCs and sensors in real time, entirely on your hardware.
One install, three built-in profiles — book a demo to drive them live.
The sensor wall with live AI analysis, top risks, risk trend and the action log. Plant composition health up top; every alert one click from its evidence.
Value delivered, downtime avoided, work orders, energy and quality — per-asset health rolled up to what it costs and what NFO saved.
Routing and system health — how signals flow to expert packs, how the reasoning mix is weighted, and how deep the system is thinking. Read-only, always.
Modbus TCP, OPC UA, or replay for commissioning. Tag maps live in plain YAML; switch gateways with a single setting.
Threshold, missing-signal, spike and drift detection; ML anomaly scoring; cross-tag correlation; episode lifecycle — all governed by one fabric.
An ongoing condition is one persistent episode with per-episode emit cooldown. Stale episodes auto-close. Severity is unambiguous: info, warning, critical.
Attach manufacturer PDFs to any sensor. Layout-aware extraction reads real spec tables, so recommendations cite rated limits, maintenance intervals and model ids.
Per-user installer with no admin prompt. Native window, system tray, single-instance lock, and one-click checksum-verified auto-updates.
Write-ahead-logged local store, indexed queries, response caching, and a single batch endpoint that returns the whole dashboard in under 3 ms.
Because NFO is 100% on-premises, there is no cloud tenant, no vendor tunnel, and no outbound AI service touching your plant. The reasoning runs beside your PLCs, not between them and the internet — so malicious-AI attacks on your controllers have nothing to grab onto.
All data, models and reasoning stay inside the building. The initial release ships fully air-gapped — NFO installs and runs with no internet connection at all.
NFO listens to Modbus and OPC UA; it never writes to the PLC. The AI can recommend — only your people can actuate. Supervised write-back is on the roadmap, opt-in and gated the same way.
Every finding and recommendation is persisted locally with its evidence, so you can verify exactly what the AI saw and said.
NFO installs like a desktop product because it is one. No IT ticket, no server provisioning, no cloud onboarding. Close the window and it keeps observing from the tray.
A pilot takes an afternoon to stand up: one machine, your PLC tags, your datasheets. Keep the findings either way.