by Julian Goldsmith

IBM’s Steve Mills calls for analytics to come on down

Nov 10, 20113 mins
IT Strategy

There are no real break-through technologies, says IBM’s Steve Mills, senior VP and group executive software and systems.

All technological developments, however impactful are built on previous innovations. Much of what seems to arrive from nowhere has actually been in development for decades.

“It’s a case of what’s possible is not always economical”, he said when we met in IBM’s London Southbank offices.

There is a paralell the most ground-breaking technological development so far, the internet, which was in development for over forty years before people started buying goods and swapping holiday snaps online.

It’s the same for IBM’s latest big push, analytics. In fact IBM isn’t the only big vendor to be talking up analytics.

Mills believes that’s no accident. He thinks we have reached a stage in the technology’s development where it is not only possible, but economical to deploy.

And the effects on people’s daily lives could be just as fundamental, if perhaps not in quite such a spectacular way. Analytics, he says was previously only affordable by the big market leaders.

Now it can be economically deployed by their smaller competitors, levelling out the playing field.

The second part of the equation for him is the explosion of cheap mobile comms and connectivity, which has enriched users’ experiences of IT in one direction and allowed a wider stream of data back to core systems in the other.

Size and location are no longer determining factors for business success or even dominance.

He said: “This allows big complex projects to operate anywhere. IT is no longer tethered to one place.”

Mills thinks that the changes that analytics will bring could be significant but subliminal. People will notice the affects, but may not know what IT developments are behind it.

He cited that fact that Boeng’s design of the 777 was entirely digital, the first time it had not resorted to building a physical model to test out design innovations.

He said: “It’s a reflection on the computing power companies have at their disposal now.”

Processing power, the third piece in the analytics puzzle, has traditionally been brought to bear on systems that required high volume transaction processing, such as trading systems.

It’s only recently that grunt power has been focused on data of a more unstructured nature. He cites the Watson project, where a computer beat human competitors at the game of Jeopardy and which is now being developed for medical diagnostics.

Mills said: “Watson is a learning system which associates attributes, in the same way that a human brain works. That gets interesting when you apply it to real-world problems.”

Watson doesn’t just record data, it sorts it and forms an understanding of what that data represents.

The implications for the usefulness of this sort of computing could be enormous.

It means industries other than data crunchers like stock exchanges, could enhance their abilities to enhance the quality of people’s lives – whether they are electricity providers, healthcare organisations or retailers.

Mills saves the most compelling example of the good use of analytics for last.

He said: “We are working with a hospital in Toronto, in the premature baby maternity ward. They are using analytics there to measure the vital signs of the babies in their care and comparing each one with the rest to get an indication of their general well being. It means nurses can be flagged if a baby is becoming particularly stressed.”

Medical staff could be saving new lives, without the knowing about the algorithms or years of development that brought them that crucial and timely information.