SensiML TestApp

Real-time, on-device testing of SensiML Knowledge Packs

Seeing is Believing. See Results Fast with TestApp.

Validating model accuracy in real-time on the intended edge IoT device is vital to building confidence your product performs as expected. The time gap between model simulation and working IoT device can take weeks or months with traditional design methods. With SensiML’s AutoML workflow culminating with device testing with TestApp, you can get to a working prototype in mere days or weeks.

Options available in Analytics Studio can provide the necessary wrapper code to provide real-time model result streaming over a variety of output interfaces including Bluetooth LE. Thus a model built in SensiML Analtyics Studio can be flashed and tested on device directly with no additional coding required.

Test or Demo? We Have a TestApp For That.

Whether testing internally to gain insight into model performance and generalization, or using TestApp to tell a story to customers, partners, management, or investors, TestApp can accommodate your needs for quick verification and test.

Available as a standalone Windows 10 PC test application and on Android mobile devices, TestApp can be quickly configured to display in real-time and log classification results from your model running on actual hardware.

SensiML Test App

  • Test tool for model validation on device
  • Real-time and logged collection
  • Rich visualizations and debug info
  • Windows and Android versions
  • Skinnable graphics for showcasing demo applications
  • Android version available as open-source reference code for user modification as required.

Validating Results

Watch how TestApp is used to validate a working model in real-time on the target device.

Results from the SensiML Knowledge Pack sensor algorithm are output over Bluetooth LE and displayed in TestApp along with the underlying feature input vector. Visualization of feature space is possible by selecting any two features to plot graphically in TestApp.

SensiML Fan Fault Demo

Watch an example of TestApp used to show a seven class fan state inference model at work.

With TestApp you can quickly show concept demos for intelligent sensing applications. For this demo, we demonstrate the power of the analytics using only 3-axis accelerometer sensor data to accurately detect a variety of fan states and faults.

SensiML Data Depot

Explore models in our growing database of example datasets and SensiML generated AutoML models and try them using TestApp.

The SensiML Data Depot provides a repository for models you can try and adapt to suit your needs.

Raw Signal Capture to Data Insight Labeling (Data Capture Lab Phase) to Algorithm Generation to Firmware Code Generation (Analytics Studio Phase) to Test, Validation and Support (Test App)

The SensiML Endpoint AI Workflow

Watch The Workflow Video