Skip to content
Back to work

AI · Manufacturing

FactoryMind

An AI-powered fault diagnosis tool that helps manufacturing engineers spot failures before they cause downtime.

Built for
Engineers on the line
AI
Fault diagnosis
Status
Pilot deployment
FactoryMind home screen: live line status timelines for Line 1 and Line 2, and the AI shift summary

Problem

Downtime that's discovered too late

On a busy production line, small faults often go unnoticed until they cascade into costly downtime. Engineers were relying on manual inspection schedules and tribal knowledge to catch problems — reactive rather than predictive.

The brief was a tool that could flag likely faults early, using existing sensor and machine data, without requiring a data science team on staff to run it.

Shift walkthrough showing line stoppages, can counts for each line and a stoppage total
The shift walkthrough: stoppages and output for each line.

Approach

Diagnosis engineers can act on

We built a diagnosis pipeline that ingests machine telemetry, applies anomaly detection tuned per asset type, and surfaces plain-language fault explanations rather than raw model output — so an engineer on the floor can act on it immediately, not just a data analyst.

The interface was designed around the reality of a factory floor: high contrast, large touch targets, and a triage view that ranks issues by likely impact on production.

Quality screen: scan raw material labels with the camera and the AI reads the fields
Quality checks: scan a label and the AI reads the fields.

Result

From reactive to predictive maintenance

FactoryMind is currently in pilot deployment. Early results point to faults being flagged well before they would previously have been caught by manual inspection, giving maintenance teams a genuine head start.

More screens

Around the app

FactoryMind sign-in screen
Sign in.
Production tracking: choose the line and product, then log cans produced each hour
Production tracking, hour by hour.
About FactoryMind: a fault knowledge base built for the people who work on the line
A fault knowledge base for the people on the line.