Data Sensors

Turn measurement data into a working sensor

Data Sensors is a measurement data viewer, a Simulink and Stateflow model viewer, and a virtual sensor workbench in one desktop app. Train a virtual sensor from real measurement data, rank it against alternative model types, and export it as standalone C/C++, Python, or MATLAB code, ready to run on an MCU, in a Simulink block, or in any other environment your own toolchain doesn't already cover. Runs fully offline, with no cloud dependency and no accounts.

See pricing Request a demo
Windows & macOS No account required Works fully offline 7‑day free trial
Data Sensors main window showing multiple signal strips and the calculated signals panel

Features

Everything needed to go from a raw measurement file to a working, exportable sensor model, all in one desktop app, with nothing sent anywhere else.

Measurement Viewer

Open CSV, MAT, MDF/MF4, or TXT files - or import a CAN bus trace - and view any number of channels across multiple strips, with independent per-strip axis grouping, dual draggable cursors, and synced zoom/pan across the whole session.

Learn more
Strip view with two cursors placed and a values table showing readings at each cursor

Simulink Viewer

Open a .mdl or .slx block diagram and any linked Stateflow chart read-only, with real block icons, real parameters, and signal tracing - no MATLAB or Simulink license required, so engineers who only need to review a model never need a seat of their own.

Learn more
Simulink model viewer showing a real .slx model's block diagram, including a Stateflow-linked block and real block icons

Calculated Signals

Derive new channels from existing ones with a live formula editor that references any real or already-calculated signal by name, catching malformed expressions as you type.

Learn more
Calculated Signals panel defining a new signal named power from the expression torque times motor_speed

Virtual Sensor Training

Train Random Forest, XGBoost, or Neural Network models on your own measurement data, whether that's a single file, several files, or a whole folder of recordings at once, even if different files name the same signal differently. Any of the three model types can auto-tune its own hyperparameters (Bayesian, random, grid, or surrogate search) instead of requiring hand-picked values.

Learn more
Create Virtual Sensor dialog with input and output channels selected, training data source, and auto-tune hyperparameters enabled

Model Competition

Train several model types concurrently on the same data, rank them by held-out validation accuracy, then export any entry from the leaderboard, not just the winner.

Learn more
Model Competition setup showing Random Forest, XGBoost, and Neural Network competitors selected for training

Standalone Code Export

Export any trained model to C, C++, Python, or MATLAB, in full precision or int8-quantized, with a live size and accuracy preview before you save, so you know exactly what you're shipping. Already have an exported C, C++, or Python model on disk? Import it back in to run it against newly loaded measurement data, no retraining required.

Learn more
Export Virtual Sensor dialog with C++ selected as the export format

Save & Share Sessions

Save a whole session, including strip layout, signals, calculated signals, virtual sensor signals, strip background color, and sample rate, as a single config file to reopen later or hand to a colleague. Save, Save As, and Open all work like any desktop app, and the window title shows an asterisk the moment anything is unsaved. Export any strip, or all of them, as a PNG for a report.

Native save dialog saving a Data Sensors session as torque_temperature.json

Licensing Without Accounts

No sign-in, no telemetry, no phone-home. A license file (or a company-run license server for teams) is the only requirement.

Learn more
License activation screen showing the machine ID and options for a license file or license server

Pricing

Every tier can view measurement data and Simulink/Stateflow models, and train and evaluate virtual sensors. Exporting a trained model to deployable code requires a Virtual Sensor license.

Trial

View & train, seven days

€0

  • 7 days, starting from first launch
  • View measurements & Simulink models
  • Train & compare virtual sensors
  • No account or credit card needed
Get the app

Viewer

Single machine, view & train

€1,000/year
excl. VAT

  • Licensed to one machine
  • Measurement & Simulink/Stateflow viewer
  • Train & compare virtual sensors
  • Free updates for the license term
  • Email support
Buy now

Virtual Sensor

Single machine, full export rights

€10,000/year
excl. VAT

  • Everything in Viewer
  • Export to C, C++, Python & MATLAB
  • Use exported code in your own products, royalty-free
  • Free updates for the license term
  • Priority email support
Buy now

Floating (Team)

Minimum 5 concurrent users

Custom/year

  • From 5 seats, shared across your team
  • Viewer or Virtual Sensor capability
  • Self-hosted license server (Docker or plain Node)
  • Contact us for volume pricing
Contact sales

Prices shown in EUR, excluding VAT where applicable. Licenses run for a 1‑year term; renewal is a new license file, no reinstall required.

Contact

Questions about pricing, a floating-license quote, or technical support? We read every message.