Anomaly Detection - for an electric drill
- A few of Rapid IoT's sensors are utilized- gyro/accel/mag
- Rapid IoT is trained to send an alert when an out of bound condition occurs
- The same principle of anomaly detection can be applicable for other verticals- e.g. temperature, humidity, air quality for Home and Building automatio
Thanks to Theophile LeRoy for putting this together
- 既読としてマーク
- 新着としてマーク
- ブックマーク
- ハイライト
- 印刷
- 不適切なコンテンツを報告
That is very cool app, and I saw a live demo on a i.MX processor at the CES 2019. Is the app available for download?
- 既読としてマーク
- 新着としてマーク
- ブックマーク
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi Sudhir, Not yet final or available- we've replaced the drill with a small DC electric motor.
Stay tuned- and what do you plan to do with the app?
- 既読としてマーク
- 新着としてマーク
- ブックマーク
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Hi Jay,
We will like to explore that tool with additional water quality and air quality sensors. We have also developed the ability to run recurrent neural network (GRU) on K64 Cortex M4, and would like to compare some of the algorithms like SVM in your tool with the GRU time-series anomaly detection.
-Sudhir
- 既読としてマーク
- 新着としてマーク
- ブックマーク
- ハイライト
- 印刷
- 不適切なコンテンツを報告
Cool!
But I have so many questions on how this is done...
The GUI shown is done in Matlab? what kind of model is trained ? Is it a neural network or something else ?
I guess the model and the anomaly detection runs on the PC, and the rapid IoT is used to stream only the data ?
- 既読としてマーク
- 新着としてマーク
- ブックマーク
- ハイライト
- 印刷
- 不適切なコンテンツを報告
GUI was done in C++ I think.
2 models can be trained: SVM (Support Vector Model) or GMM (Gaussian Mixture Model). There is no NN in this demo.
The anomaly detection fully runs on Rapiot Iot (training include) and a serial bus streams data to the PC, which is only used as a way to display and configure the parameters easily.
We are finishing another version of the program which does not need any PC - demo video will be posted soon... during Embedded World showcase.