MeOn – IoT Industrial monitoring – Slush

In a 2-man team, we have co-founded the project MeOn which uses the SaaS model in remote monitoring systems for industrial sensors and forecasts machine failures.

MeOn was in the top 3 finalist in the Elisa IoT Innovation Challenge 2016 and I had the privilege to present it in Slush 2016. Check out the video presentation.

MeOn is a noninvasive solution to determine the health of industrial motors and forecast failures by sampling the vibration of the machine and interpreting the data.

The compact yet powerful sensor is able to sample and analyze the vibrations of the equipment, and to send the data directly to the cloud.

On the cloud there are smart algorithms able to understand and learn from the machine to forecast when a failure is going to happen.

I did the next implementations/integrations for the prototype:

Cloud side:

• Gateway built in node.js to translate messages coming from the MeOn nodes to Elisa – IoT cloud platform: Thingworx
• Deployment of the gateway in a Linux instance deployed in Amazon Web Services
• Configuration and usage of Elisa IoT cloud platform: Thingworx
• Setup of 3rd party plugins, Twilio, to send SMS notifications when alarms are detected

Embedded devices:

• Built signal conditioning circuit to process vibration coming from a piezo sensor
• Integrate digital accelerometers through I2C protocol
• Test of different IoT nodes. Programming of most of them was done in C++ like languages
• Integration and modifications to a Fast Fourier Transform algorithm to process the vibration data

Artificial intelligence:

• Spectral analysis of signal coming from the node
• Anomaly detection algorithm using a modification of k-nearest neighbor algorithm

MeOn benefits

MeOn indicates its status through color codes (blue: connected to the cloud)

Elisa Press release of MeOn – The video shows the first prototype, which later evolved to the design shown in the previous diagrams

(Click image to go to the article)