Big data to predict equipment failure? IBM says yes

Big data to predict equipment failure? IBM says yes

Summary: It's all about avoiding costly supply chain interruptions.


This morning, IBM unveiled a new software and service package within its Predictive Asset Optimization product that is intended to predict and prevent equipment failure.

The idea: use big data, that nebulous concept, to identify irregularities (and calculate the potential risk of them) in instrumented assets so as to avoid slowdowns or failures in the manufacturing process.

Hardly a sexy topic, but the ramifications are massive. If there's a slowdown (or outright stoppage) in the supply chain, the losses pile up immediately. That's the downside to being extremely fast and efficient: that speed can come back to bite you when the operation goes off the rails.

(See: Crisis, Financial; 2008 edition.)

The potential damage, of course, depends on the product in question. You can imagine how awful a supply chain interruption would be to a company -- say, a large grocer -- that produces perishable items. One slowdown, and you've got pallets of food spoiling in your warehouse, and millions of dollars in potential revenue evaporating every minute.

IBM thinks it can improve uptime by deploying sensors on the assets of a production line and analyzing the liquid gold -- er, data -- that flows forth. By applying its analytical insight to the numbers, it says it can not just avoid catastrophic failure, but predict it, too -- important not only to avoid trouble but also avoid the additional costs; three to 10 times as much, IBM says -- that emergency maintenance carries.

The market opportunity for the company is real: roads, bridges, water supply systems, sewers, electrical grids and telecom networks are all in its sights. It's not just a run at factories, and spans some of the global economy's largest sectors, from automotive to electronics, transportation to telecom. For some, it's catastrophe avoidance; for others, it's customer support savings (e.g. product warranties). 

The new capabilities will be offered through IBM's new Advanced Analytics Center in Columbus, Ohio.

Topics: Big Data, Enterprise Software, IBM

Andrew Nusca

About Andrew Nusca

Andrew Nusca is a former writer-editor for ZDNet and contributor to CNET. During his tenure, he was the editor of SmartPlanet, ZDNet's sister site about innovation.

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.


1 comment
Log in or register to join the discussion
  • Not new technology

    A decade ago, I interviewed for a five-year old company that designed and implemented predictive failure technology for equipment ranging from aircraft engines to locomotives to nuclear power plants. They were able to show, for example, that scheduling a replacement part on an aircraft engine during a scheduled maintenance window could save an airline as much as $1 million if the same jet was pulled from service in response to the actual failure. This was a small, growing company and the engineers all seemed quite brilliant...too bad I couldn't wait for their next big contract that would have given them the budget to hire me, it would have been a fascinating place to work.

    What this technology does in essence is create an emissions fingerprint for the equipment being monitored. This includes such things as normal vibrations, sound, and even the chemical make up of combustion exhaust among other things. The idea is provide a snapshot of a "healthy" system and when these indicators change, further analysis regarding the root cause of the change begins (and my recollection of the technology ends). The company I interview for provided both client-side monitoring (the client kept their own servers and monitored the data) as well as hosting servers and providing analytics at their own facility.

    I imagine with their resources, IBM could make the technology available for less complex equipment, such as line equipment in plants or even trucks used to deliver product. Custom engineered solutions like the company I interviewed with would be prohibitively expensive to implement for less-than-high-value equipment, but with a larger infrastructure, IBM could probably provide more rapid deployment more companies. It's cool technology, but not very new.