As I continue to reflect on the talk I gave on What Datacenter Operators Can Learn From Rolls Royce (included at the end of this post) at Datacenter Dynamics New York, I realize that the timing of the topic was interesting based on the continued search for Malaysian Airlines MH370 where it was Rolls Royce who announced 3 days after my presentation that MH370 was flying for hours after it disappeared from radar as they were monitoring MH370’s jet engines during flight.
The main premise of the presentation is that airlines long ago realized they could greatly benefit from working with Rolls Royce to monitor real time the state of health and key performance data of their jet engines in their aircraft to improve safety and reliability. By working with their trusted advisor and partner, the airline industry migrated to conditions-based maintenance from time-based maintenance which also extends the lifecycle of those jet engines. Datacenter operators can do the same thing! Like jet engines being one of the key vulnerable components of an aircraft, the power continuity infrastructure of datacenters (UPS Systems, Battery Systems, Power Distribution Equipment, Switchgear, Generators) is the most vulnerable part of the datacenter per ongoing studies from The Ponemon Institute.
Many of the leading operators like Equinix, CenturyLink, Computer Sciences Corporation, RagingWire, Digital Realty Trust, Cologix, ViaWest, Charles Schwab already are working with Canara just like the airlines work with Rolls Royce to eliminate risk and save money while doing it. Canara’s formula for our success with these operators is nearly identical to the Rolls Royce process:
Sense – Canara deploys many sensors in the power infrastructure of the datacenter to gather key information about each unit of power backup (battery) or power distribution (circuit). This can include things like the voltage, ohmic, AC ripple on DC systems, temperatures, current measurements of circuits and other key measurements.
Acquire – All of that real-time “big data” has to be consolidated to a central location in order to make it useful. This includes acquiring data from non-Canara made equipment.
Transfer – Canara transfers the data to our Monitoring Operations Center like Rolls Royce transfers jet engine performance data to their Bristol, England Monitoring Operations Center.
Analyze – Canara’s time proven and powerful cloud based analytics engine takes all that big data and runs ever improving heuristic algorithms and prognostic algorithms based on 20 years of data analysis in this area (over 2 billion data points) and automatically flags anomolies to investigate. These flagged anomalies are then analyzed by real people who are experts in their area to then render a recommendation of whether to ignore, continue to trend or Act. And because we are doing this every day, these actions tend to be conditions based maintenance oriented to pre-empt a real problem before it actually occurs as opposed to an urgent alarm state.
Act – When the recommendation is to act, Rolls Royce has the maintenance crew at the ready before the plane even lands with the specific plan of action already created. In the datacenter world, Canara’s time sensitive and recommendations include replacing a battery or entire battery string on a UPS or Generator or DC Plant, replacing a UPS capacitor, adjusting buss voltage, adjusting temperatures, adding circuits for a customer, investigating a hot spot in power distribution equipment and many others.
The parallels of jet engine management to datacenter power infrastructure management are really interesting the deeper you get into it. I plan on doing a Part 2 of What Data Centers Can Learn From Rolls Royce in the coming months to further explore the parallels. I hope you enjoy part 1…