See the loss coming.
Predict deviations before they cost you.
TeepTrak's Machine Learning layer learns what normal looks like on your lines and flags anomalies before they turn into downtime or scrap — surfacing patterns no human has time to find.
app.teeptrak.com/perftrak · Machining line 1
LIVE
Cause qualified by operator: tool change · 4m 12s
Machine Learning runs in plants of
The signal is already in your data. Machine Learning finds it before you feel it.
It learns your normal
The model studies each line's real rhythm — cycle times, stops, parameters and quality — and builds a living baseline of what good actually looks like.
It flags the abnormal early
Drift, creeping micro-stops, an emerging quality pattern — the model raises it while there is still time to act, not in next month's report.
With the likely cause attached
A live Pareto of losses by line, shift and cause. Your morning meeting starts from the same facts — and your CI team attacks the biggest loss first.
How Machine Learning works
Three steps from your existing data to an early warning you can act on.
Feed it your data
Machine Learning runs on top of your existing TeepTrak data — OEE, stops, cycles, parameters and quality. Nothing new to install.
Operators qualify causes
When the machine stops, the tablet asks why. The operator picks from your loss tree — two taps and production continues.
2 taps, zero typingSupervise in real time
Consolidated live dashboards: OEE by line, Pareto of causes, shift reports, smartphone alerts, raw-data APIs for your BI.
Machine → plant → groupThe supervision platform, exactly as your team will see it
Every line, live, on one screen
Walk into the workshop — or open a browser anywhere — and see machine status, OEE and current losses update in real time.
● 1 stopped — cause: misfeed
The Pareto that builds itself
Every qualified stop feeds the loss tree. By Friday you know your top three losses for the week — by line, shift, product or team.
Shift reports, written automatically
Production counts, OEE, top losses and scrap — compiled and sent at end of shift. No more Sunday-night Excel.
Know the moment a line goes down
Configurable alerts to any smartphone — machine stopped, OEE under threshold, changeover over target. React in minutes, not at the morning meeting.
14:32 · running 0 min · no cause yet
14:18 · 22 min vs 15 min target
13:55 · OEE 92% · shift on track
Insight, not a black box.
Adoption is where OEE projects die. PerfTrak's tablet interface was designed with operators, for operators — no typing, no menus, no blame. Just "why did we stop?" and back to work.
Installed in five steps — without stopping the machine
Install the module
In the electrical cabinet, or outdoors with external boxes
Pick your signal
0–24V PLC signal, external sensor or OPC UA
Position the tablet
At the workstation (or go tablet-free with Light / Live)
Configure
Products, standard times and your loss-cause tree
Start tracking
Production deviations captured live, from minute one
What could early warning save you?
Drag the sliders. Estimate based on a conservative +8 OEE-point gain — PerfTrak customers average +10 to +15.
per year, by catching drift early
≈ 6,400 machine-hours back in production
Estimate my upside →Choose your Machine Learning
PerfTrak
Anomaly detection on your full TeepTrak dataset — drift, micro-stops and quality patterns flagged across every line.
PerfTrak Light
Start with one use case — downtime or quality — and expand as the model proves its value.
PerfTrak OPC UA
Push predictions into your BI, MES or maintenance system via API — Machine Learning watches, your systems act.
The results our customers see every day
"Quickly adopted by all our production lines thanks to its intuitive interface and clear, real-time dashboards. Operators use it daily for short-interval control."
"Real-time OEE tracking and stoppage capture at 3 sites. Fast deployment, operator-friendly interface, clear loss trees, actionable alerts, easy BI exports and APIs."
"Deployed on all our automated equipment. Strong adoption by operators, elimination of paper and Excel, rapid OEE gains on micro-stops and team organization."
Your data, wherever you need it
Standard APIs and raw-data access. Feed your BI, your ERP, your data lake.
Frequently asked questions
See what your data already knows — on your lines
A 30-minute working session with a TeepTrak engineer. We'll map your loss profile and show you PerfTrak running on data like yours.