Mark Grabb, Technical lead for the global analytics practice within GE, took some time at Big Data NYC to discuss GE’s vision for the Industrial Internet. GE is leading the charge in delivering what industrial customers want: zero unplanned downtime, highly reliable, intelligent assets that are sensor-enabled, and machines that use data and analytics to be economically efficient within their own ecosystem.
Since GE’s business is in Industrial equipment like gas turbines, airline engines, and other highly sophisticated machinery, the analytics is made more complex with the inclusion of physics, material science, and thermodynamics that must feed into it along with traditional data analysis. Yet these analytical capabilities have reduced the need for brute force investigation of issues that once required the time to take a machine completely offline. By collecting imaging and sensor data all along the lifecycle of the asset, engineers can better infer when to insert preventative maintenance cycles, as well as feed findings into better designs, creating a virtuous cycle.
In heavily regulated industries such as industrial manufacturing, there are increased requirements determining how information can be stored and accessed. As the underlying technology behind its Predix platform, Pivotal helped to jump start the capturing and analysis of data with industrial strength solutions. Grabb states, “The flexibility from an analytics perspective of having accessibility of all the data is driving so much analytics and speed. We would spend 80% of time getting data in order, and the cool stuff I’m so excited about was only 20% of the time. Now what we’re seeing dramatically shift is all that exciting stuff is turning into 80% or 90% of the time. […] You cannot afford to have time messing around to get your data in order. That is why the partnership with Pivotal has been vital to us. “
GE has enabled a data lake strategy in which all data of interest in analysis is housed in a repository commonly accessible to data scientists and business analysts alike. This strategy has allowed for large bodies of sensor data, which in the past were only sampled, to be captured and analyzed in their entirety. The Pivotal platform has enabled flexibility in analysis: at times you need to create analytics that move quickly, get down to the data—Pivotal has allowed the Predix platform to do this quickly. Other times, the analytics are so complex that map reduce is not possible and data needs to be brought up: Pivotal has industrial strength solutions for this. This has jump-started GE’s approach to big data.
About the Author