We worked on a similar project in collaboration with Los Alamos Labs. In particular to do reliability and failure prediction for systems that were composed of elements that only rarely failed. So called 'Black Swans'. This was eventually licensed out via a third party. Examining the difference between the ideas. Will report back here with more information. A related project looked at out of stock condition on a store shelf as a failure that could be predicted by multiple sensory inputs.
Pointer to some of the above work in R&D Magazine.
This was eventually offered for use by KPMG Consulting. in 2005.
In FastCompany:
GE wants to give industrial machines their own social network with Predix Cloud ... GE is selling a new service that promises to predict when a machine will break down, so technicians can preemptively fix it. .... "
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment