Inspiration - The Quest to solve industrial problems and leveraging Predix Platform to do the same motivated us to conceptualize and ideate for one of the most mobile and important persona in Industrial Businesses - Field Service Engineers
What it does - Business Problem: About 60Bn of Services Backlog and operations cost and productivity is a big play today.
Solution: Enable the FS engineer with Real time data and insights for Optimization of the Model through Digital Twin and also enabling them to initiate part repair through Mixed Reality on a Hololens. Gain: Target to get atleast 1% overall productivity benefit for GE’s 60Bn Services business and for a world industrial Services economy of $11.9 Trillion
How we built it : As part of the solution, we have a model prototype of a Gas Engine with 4 Piston rods and we have connected Vibration and Noise sensors onto the same and pushing that data into time series database. From there, we are running Analytics using algorithms for Expected behavior vs actual behavior for optimization of the asset and also, we have built an algorithm for Workforce analytics to identify the right skill for the right service request using a SMS Notification. Once the FS receives the notification, he wears a Hololens device where we have coded an app on the device for the FS engineers to see the Digital Twin of the Turbine in 3D and expand the turbine, see the heat map overlay of the parts and parts that are due for service. Once that is done, he initiates part repair through Mixed reality
Challenges we ran into - One of the challenges that we ran into was around the Pieze vibration sensor that we had was not sensitive enough for the malfunction that we were trying to impersonate on the model prototype, so we had manually calibrate the sensor (hard) and mitigate that challenge. Also, we wanted to have a Rotation sensor in place but the Grove kit didn't come with the Rotation sensor.
Accomplishments that we're proud of: The biggest USP of our solution in terms of what we were able to accomplish is the usage of Predix Components and Services as below...
- Predix Machine - Vibration sensors and Noise sensors
- Used 2 Predix Catalog Services - Anomalies, Run time aggregation
- Predix Asset Model
- Used 7 Predix Services - (Postgres, UAA, Analytics Catalog, Analytics Run Time, Analytics UI, Time series, Asset database, PX)
- Digital Twin with Mixed Reality on Hololens and iPAD
- Predix Analytics - Time of Expected Arrival Workforce Analytics, 2 Algorithms contributed to Platform