IBM Watson is a project receiving the highest levels of support and encouragement from IBM, having got its big launch at the start of 2014 with the formal set-up of the IBM Watson Group.
At the time, IBM CEO Ginni Rometty characterised this launch as a major turning point for IBM. It was, she said, as significant for the company as the System/360 or the IBM PC before it. The very name, taken from IBM founder's son and 1960s mainframe champion Thomas J Watson Jr. hints at the level of company commitment involved in a cognition technology it uses.
Watson started as a supercomputer. Its role was largely unexplained and made all the more confusing by the regular chess-playing computers IBM used to roll out from time to time. The first public outing and test for the putative 'cognition' technology that Watson offered was in the very US-centric environment – in 2011 Watson appeared as a competitor on TV quiz show Jeopardy (where it trounced the humanoid opposition). More echoes of IBM brother Deep Blue and Garry Kasparov facing each other off across an electronic chess board.
But this was no game-playing robot. Taking part in a Primetime US TV show may have been a PR stunt that made massive sense back at IBM's Armonk, New York HQ, but since most potential buyers of this technology outside the US don't even know what Jeopardy is, it may not have been a particularly smart one globally.
Rometty claims Watson moves IBM to the third 'cognitive' computing era. Watson is able to manage, understand, learn and hopefully derive usable information from large amounts of data. However, for some critical IBM observers, this remains a technology in search of an application. That seems unnecessarily harsh: There is no question that we are already swamped in Big Data (2.5 billion gigabytes a day is being generated apparently), so any tools that offer to make sense of part of that huge data set will undoubtedly have utility.
What may be true and open to further exploration is the difficulty of the undertaking. The solutions and applications which require this kind of processing are also typically big, difficult, ones, and those that have not always yielded success in the past. Which make this is a particularly hard area for development and a real challenge for any startup (IBM's own description of their 1,000 person Watson Group) to create and manage. It is tough and slow job to sell the real world applications and value of the cognitive data learning and analysis tools Watson represents as part of the big ticket project it will inevitably be. But sales is what made IBM great in the first place.
Healthcare has been one of IBM's early targets. Recent announcements in the US show continuing work in this sector: Explorys and Phytel will become part of the Watson Health Cloud. Explorys has compiled a large global healthcare database, built with financial, operational, and medical record source systems. Phytel works with US healthcare providers' current electronic medical record technologies to reduce patient hospital readmissions and to automate and improve patient communication.
In fact, right from the start, eighteen months ago, we learned of work being done in target high profile areas of healthcare with Watson, such as Oncolgy, to assimilate, analyse and provide decision support tools to medics working with cancer patients. Each one of us generates a vast amount of medical data over our lifetime. So it's small wonder that stretched doctors find it hard to even reference, never mind use, this data to analyse and guide the clinical decisions they make on our behalf.
Which is why automating and sharing patient records is just a first step in a bigger journey. The NHS has a particularly painful recent history on patient records. Since the demise of the NHS's huge and ultimately doomed National Programme for IT (NPfIT), CIOs for individual health areas have been looking at similar projects for themselves; health area by area. Some of the solutions, such as US vendor Epic's healthcare records system, are the very same companies also now buying into the IBM Watson technology, so that they can offer better ways of both delivering the data and offering meaningful analysis about medical causes and clinical treatment based on that data.
Cambridge University Hospitals NHS Foundation Trust (which includes Addenbrooks Hospital) is the first UK health authority to have bought the Epic system. The Royal Devon and Exeter NHS Foundation Trust has also confirmed that it is planning to implement the same new system, and Papworth Hospital NHS Foundation Trust has also committed. Other health authority CIOs with similar requirements will be watching closely to see how these high profile IT projects work out n the coming year.
IBM knows it has to play the long game to win over CIOs to Watson's abilities. Healthcare is just one area of business that it would like to be active in, but it is one where it seems to be making some early progress.