Industry Solutions
Industry Solutions
See what changes when AI agents work on shop-floor systems, shown on a manufacturing MES and a logistics WMS that actually run
Both are implementations for demonstration and learning. They do not present production track records or performance figures.
See It in the Factory and the Warehouse
For an agent to do real work, it must read the facts on the floor, get approval, and check the result afterward. Both solutions show that whole path from start to finish.
uEngine MES
A manufacturing execution system that records every change as an event. It calculates the impact of equipment failures, ties several sources together with a semantic layer, and has agents execute on a person's decision.
uEngine WMS
A sample app covering stock shortages, quotes, approvals, receiving, inspection, putaway and warehouse equipment control. It connects to Process GPT agents through MCP tools, and its source is public so you can install it yourself.
Same Structure, Different Floors
Both solutions are layered the same way. The shop-floor system stays as it is, with tools agents can use and human approval placed on top.
| Layer | What it does | uEngine MES | uEngine WMS |
|---|---|---|---|
| Shop-floor system | Records the facts of the work | Event-sourced MES. 10 services, 145 screens | Supabase WMS. 17 screens, 37 tables |
| Shared meaning | Ties several systems into one judgment | Semantic layer. 7 sources, 5 judgment concepts | Not needed, since it is a single app |
| Agent tools | Defines what agents can do | read·plan·exec function catalog | 83 MCP tools |
| Process and approval | Sets the order of people and agents | Process GPT response process and decision card | Process GPT replenishment process and purchase approval |
| Records and oversight | Keeps what happened and why | Event store, execution ledger, rule versions | Agent Decisions, Audit Log |
Have more questions?
Want to know more about Industry Solutions or have a question? Get in touch anytime.