InduQT

InduQT troubleshoots industrial equipment from the evidence a plant already has: live signals, alarms, equipment manuals and maintenance history. It covers boilers and burners, VFDs and drives, CNC machines, servo axes, compressors, pumps, chillers and conveyors.

With the connector, Claude can connect a plant's data from the conversation: browse an OPC UA server and preview an MQTT broker before connecting them, create a site-scoped ingest key with a ready Node-RED flow for gateways, historians and proprietary systems, load Excel or CSV history from a link, and add manuals, photos of alarm screens and work-order exports. InduQT reads every file, scanned pages included, and indexes it for retrieval.

Once data flows, InduQT's decision engine watches every signal against its normal range. When something leaves it, the engine profiles the machine's recent history, finds which signal moved first and in which direction, separates background changes from the event, scores the causes it knows for that kind of equipment, and opens an investigation. Claude can then start or update investigations: the diagnosis names the likely cause, the evidence with times, the first actions and the single check that separates the leading causes, citing the machine's own manual pages and the team's past work orders. Technicians add what they find and the diagnosis updates.

New users can try it immediately on InduQT's public live demo (a boiler on a public OPC UA server, Helsinki trams over MQTT and a NOAA tide gauge) and search the troubleshooting knowledge base for fault codes without connecting anything. Assistants connected through OAuth cannot delete data. AI diagnoses count toward the user's InduQT plan and the connector says so before running one.