In a mining scenario, heavy trucks/vehicles are deployed to bring the ore/mined raw materials to the factory for processing.Vehicles are lined up in series and one vehicle stopped means that the next vehicle will also be stranded.The rough terrain causes frequent breakdown in vehicles which results in losses. We are enabling stabilized maintenance and reducing unplanned down times .
What it does
It absorbs sensor data from vehicle components and offers predictions on vehicle health on the go. The app has been made to talk to other eco-systems such as R and Power BI for end user consumption.
How we built it
We have developed a python based app in Predix leveraging simulated IOT sensor data from mining trucks and using survival analysis models to forecast vehicle health. We leveraged R to connect to the app and broadcast the data to Power BI for end user visualization.
Challenges we ran into
Memory and port constraints prevented us from pushing the application in predix cloud
Accomplishments that we're proud of
We have not only built an app using Predix but also integrated it with other eco-systems like R and Power BI for end user consumption.
What we learned
We learnt the power of predix, ease of use and how to use the strengths of predix to create business value.
What's next for Vehicle Health and Preventive Maintenance on Predix
We can get realtime data to fit into our data model with minimal intervention and deployed on the edge for machine to machine interaction