On-board 4-th Gen AI Anomaly Detection Device Now Available

The Onboard, real-time anomaly detection, 4-th generation AI device (CAHMS – Complexity And Health Monitoring System) has been been introduced recently (read here) and is now available for testing.

Installation is very simple. and the process is as follows:

STEP 1

Off-line analysis of client’s data sample allows to identify a suitable combination of the following parameters:

Window size, i.e., number of data samples to be processed at each step, e.g. 20, 50, 100, etc.
Analysis frequency, i.e. a time difference between steps, e.g. once every second, every 10 seconds, once a minute, etc.

These are the only two parameters that need to be defined and introduced into a configuration file on the device.

STEP 2

The following step is to create a local network connection between the CAHMS device and the data source, enabling two-way data transfer. This is a very simple procedure. The CAHMS receives data via a USB port via a USB-LAN connector.

STEP 3

During functioning, the data source computer needs to extract data windows of the previously specified size from memory, or from a CAN bus, and to write it to a shared directory on the CAHMS device with the above mentioned frequency.

STEP 4

Switch on the CAHMS. The system will start automatically and with the above specified frequency will seek for a new data frame in the mentioned shared directory.

In the same directory, CAHMS will write relevant information as well as alarm messages, if any. These are written in a format that may easily be read and displayed on client’s dashboard. Examples of outputs are illustrated in the scheme below: