Know exactly which units are affected. Then prove it.

Yes, and that is a design constraint rather than a feature. A station writes every event to durable local storage before attempting anything over the network, and replays safely when the edge returns. Nothing in the runtime path waits on a remote call — not licensing, not billing, not updates. A factory LAN with no internet at all is a supported configuration.

No. Trace reads from what you already have where it can, and captures manually where it cannot. A station with no instrument captures the same points against the same limits with the same audit fields — only the source flag differs. Plants are never uniformly instrumented, and pretending otherwise is how these projects fail.

Old is usually fine. Serial, Modbus, a TCP line protocol or a file drop all work, and many older controllers speak one of them. Where a controller genuinely cannot be read, that station captures manually and the record is still complete. We will tell you honestly at the walkthrough which of your machines can be read and which cannot.

Yes — that is the intended path. Your template is loaded as-is and versioned with an effective date, so template changes are auditable. Trace generates ZPL natively and writes it to the printer over TCP, with no print driver on the terminal, which is also what lets a printer be swapped without touching an operator station.

One line, typically two to three weeks from walkthrough to live, with close to zero production downtime at changeover. Widening to more lines is configuration. We would rather you start on one line and judge the record it produces than commit the whole plant on a promise.

Quoted per plant, not per seat. Price is driven by the number of stations, how many of them need instrument integration rather than manual capture, and whether you need direct part marking. There is no per-user cloud charge. Ask for a quote and you will get a number, not a discovery process.

Not today, and we will not claim otherwise. Trace holds the job card reference your ERP already uses, so the join has somewhere to land, but there is no built ERP connector. Data is exportable and the database is yours to query. If ERP posting is essential to your decision, say so early.

A station's spooled events are on its local disk; a replacement station picks up from the edge and no captured record is lost once it has been acknowledged. The edge box holds the database, so that is the machine to back up — and it is an ordinary PostgreSQL backup, on your schedule, to your storage.

You do, on your hardware, in a standard PostgreSQL database you can back up, query and export. There is no hosted service in the runtime path and no clause anywhere that makes your production record dependent on us continuing to exist.

Production capture continues. A lapsed subscription restricts configuration changes and cloud sync — it does not and cannot stop capture, because halting it would destroy the record for units physically on the line, and those units then could not ship at all. There is no state in the software that stops production. Ask any other vendor to put that in writing.

That is the recommended way in. The schema is multi-plant from the first migration, so growing later is configuration rather than a migration project, and starting small costs you nothing structurally.

No. Neither the software nor the company holds a certification, and any vendor telling you their software makes you compliant is selling you something. Trace is built to produce the evidence an auditor asks for — per-unit genealogy, captured values against defined limits, operator identity, and proof records have not been altered. The quality system around it stays yours.