ProSense One

The data foundation for process analysis. 16 channels ride the line and record what the part actually experiences — temperature, humidity, air velocity, zone by zone. Stream it live. Export it. Run the analysis.

Request a Test Measurement See the AI Analysis ↓
DRAG TO ROTATE

Data worth training on.

16
SENSOR CHANNELS, ONE SYNCHRONIZED RUN
14
FAULT CLASSES, PHYSICALLY MODELLED
40.8
MIN CURE HOLD, WORST PART — SPEC: 35

Measure. See. Export.

01Every channel, one record

16 channels in one synchronized dataset: part temperature on the rack, booth temperature and humidity, air velocity. What the part experienced — not what a wall sensor guessed.

16 sensor connectors on the ProSense One
02Live while the shift runs

The device is auto-discovered on the network and streams every channel live to the dashboard. Drift shows up during the shift, not in next week's meeting.

Live temperature curves from a real production shift
03One-click export, analysis-ready

Per-channel CSV, one run or many, packed as ZIP. The same files feed Excel, your ERP — and the analysis pipeline that turns a raw run into a shift report.

Control side with start/stop buttons and USB-C
AI ANALYSIS

The measurement is step one. The analysis is why.

A recorded run is not a folder of curves. It is input. We train fault-detection models on physically modelled paint-line data, pair them with an Ishikawa cause catalog built from real coating defect cases, and return a shift assessment report your quality manager can act on.

01
14 fault classes, physically modelled

Cold clearcoat oven, conveyor stall, stuck sensor: a 14-class taxonomy modelled on a real 3-layer paint line. The models train on physically modelled runs, so real measured runs validate them — the model exists before the fault does.

02
Machine fault or measurement fault?

Separating process faults from sensor faults is the models' primary target. It is the first question the analysis asks — the same question every good root-cause workshop starts with.

03
A cause catalog, not a blank whiteboard

Root-cause work starts from a prefilled Ishikawa catalog for common paint defects on plastic exterior parts — six cause categories, built from real defect cases — instead of starting from zero.

04
Verified on a running line, per part

In a full-shift customer assessment, cure was judged by the coldest part position, not the average: 40.8 minutes above 80 °C against a 35-minute spec. 5.8 minutes of margin — a number an average would have hidden.

01

Measure.

ProSense One rides the line and records what the part experiences, zone by zone.

02

See live.

Auto-discovered on the network, every channel streaming to the live dashboard during the shift.

03

Export & analyze.

One-click CSV feeds the analysis pipeline and comes back as a shift assessment report.

Automotive paint line with coating robot and IR dryers
The cure result on this page was measured on a running paint line, during a real production shift.

Specifications

PARAMETERRANGEACCURACY
Temperature−20 – 280 °C±0.15 °C (Class A)
Humidity0 – 100 % RH±1.8 % RH
Airflow velocity0 – 7 m/s±5 %
VOCConfigurablePer requirements
Particle counter IN DEVELOPMENT0.5 – 25 µmISO 21501-4
16 CHANNELS · 1 HZ IP67 WLAN + USB-C 32 GB SD — CSV 299 × 173 × 92 MM VDA 24364 LABS-CONFORM

Your line already produces the data. Start recording it.

Request a Test Measurement