
Factory automation & AI quality
Live OEE and AI quality on the shop floor
Sensors on existing machines, a camera that checks every part, and a plant dashboard on the owner’s phone.
// before & after
Same business. Different day.
Drag the handle to see how a normal day looked before, and how it looks now.
Shift log — Line 2 (night)
Machine 12 stopped ~2 hrs? (no one noted time)
Rejection: "some burr issue" — qty not counted
OEE ?? (calculated at month end)
Bearing noise on CNC-07 — check later
Customer complaint: 1,800 PPM — again!
Owner asking report by morning
The challenge
Ardent machines brackets and housings for two large auto OEMs. Rejections from the customer were rising, and every rejection meant a penalty and a phone call nobody wanted.
Downtime was logged on paper at the end of the shift, if at all. The owner found out how the month went when the month was over.
How we approached it
We did not replace a single machine. We added low-cost sensors and gateways to the existing CNCs and presses, and fixed one AI camera at the end of Line 2, where most defects escaped.
Everything flows into one live dashboard: OEE by machine and shift, downtime reasons, and every rejected part with its photo.
What we built
Machine connectivity
Sensors and gateways on 23 legacy machines, no PLC replacement needed.
AI visual inspection
A camera checks every part for burrs, dents and missing features in under 200 ms.
Live OEE dashboard
Availability, performance and quality by line, machine and shift.
Downtime capture
Operators tag the reason with one tap; patterns show up within a week.
Predictive maintenance
Vibration and temperature trends flag bearings and spindles before they fail.
// live dashboard replay
See it working.
A replay of the system we built, running in your browser with anonymised data. Go ahead, touch it.
“For the first time I can see my plant from my phone. Rejections at the customer end have dropped sharply, and the system paid for itself within a year.”
