This investment does not include the investment Big Blue has also made in the acquisition of The Weather Company, even though this will be become a major contributor to, and user of, the IoT-related data available for use with IBM's cognitive tools available in Watson.
This marks a fascinating development from IBM, for the company is not pitching to compete directly with the likes of GE Software and Hitachi Data Systems in the business of managing industrial or civil systems. Instead it is targeting the complementary role of cognitively processing the vast amounts of data IoT systems produce and infer actions and solutions that need to be made by users in a timely fashion.
According to John Kelly, SVP, Solutions Portfolio and Research at IBM, IoT is expected to produce much of the estimated 44 Zettabytes of data expected to be generated each year by 2020. The goal is to be able to use this to create opportunities for humans and machines to work together. It is not an easy task, however.
The key thing is not just that the amount of data coming out of IoT will swamp the amount of data that humans are creating, but the fact that it will largely be unstructured and very noisy. The ability to make sense of it is a huge challenge that needs cognitive management. The value comes from extracting insights from all that information using systems that learn and interact with humans in a different, synergistic way.
IBM is already well into commercialising Watson and developing new markets that exploit it. But so far other development efforts, such as Watson Health, have been based in the USA. So the move to locate Watson IoT in a large office complex in Munich is a first for the company. According to Kelly the choice is based on the city's unique place as a major industrial base in one of the most active economies in the world, with good Government support and understanding.
The new commander
The launch of Watson IoT has prompted the company to look outside for its first General Manager, and has opted for Harriet Green, who has joined IBM from her role as head of Thomas Cook.
We are talking about 9 billion connected devices now, growing to 200 billion by 2020. The currency of IoT is data but so much of it is going to waste. two-thirds is lost within seconds at the moment, but there is value in that data. And the value is in combining it with other data such as social media, news feeds and the weather. This requires a robust cloud and a platform handling 26 billion inquiries a day is the goal. It also needs to be guided by experts that understand the contexts in which the data exists, so we will be tapping into IBM's 2,000 consulting professionals around the world.
Watson is the first and only complete cognitive platform available and can reason and hypothesise, not just push out spreadsheet data. It makes recommendations on better decisions.
The basis of Watson's application to IoT is based on four core capabilities. The use of machine learning gives the primary function of learning the correct response to specific data received - a task which requires a large volume of historical data to get right. Text analytics is based on the semantic analysis of written documents, from official service manuals through to social media comments, while video and image analytics allows the tools to observe and monitor physical processes . The final component is natural language processing, which allows users to query services in their own tongue and style.
She sees Watson as giving IoT users a deeper level of human engagement, together with the ability to extend their expertise.
Many see big data and IoT as the domain of experts only, but Watson helps give many people the level of expertise they need.
She also sees it infusing products and services with cognition, using cognitive processes to help improve operations and decision making, and enhancing exploration and discovery.
A little something for the partners
One potentially important part of the Watson IoT story not told at the official launch was what servicesand capabilities will be available with the SMB community specifically in mind. This was a market sector conspicuous by its absence, though the very nature of of Watson's delivery method - as a cloud service - suggests that SMBs should be able to exploit it.
Another clue suggests that SMBs are a `soon but not yet' target, as the hint was made that as the staff grows at Watson IoT Towers, and becomes embedded in their tasks, teams will emerge that are focused on servicing the needs of specific market sectors. This suggests that pre-packaged solutions to common IoT issues might evolve over time, becoming bite-sized solutions that might well suit many of the needs of the SMB sector.
Also, the talk of openness to partnering should leave the way open to systems integrators and other niche market specialists to emerge as a channel for Watson-based solutions to the MSB sector.
Some of the existing partners were on hand to give examples where Watson will find work in the IoT market. These included The Weather Company, which was a partner until becoming an official part of IBM itself, as CEO, David Kenny, observed:
We courted for a year as partners before we married IBM. We have been using IoT a lot to improve the our mapping of the weather, particularly as the trouble with weather is that it keeps moving and changing. So the mapping needs to be done continuously. This uses vast amounts of data, including lots from commercial aircraft and the airlines.
The company already works with 200 airlines, collecting weather data from commercial aircraft as they fly through it, and providing weather data for pilots to give them advanced notice of conditions such as turbulence, which is important for issues such as the passenger experience of flying. The company also works with the insurance industry - of which Munich is considered to be the `home' city.
The retail trade uses cognitive tools to determine customer shopping decisions, where changes in the weather can often affect buying patterns. It also works with Government agencies on issues of safety, particularly in countries where bad weather `events' have a high and regular impact.
Hurricane prediction is the best example of this, where precise tracking can lead to better, more manageable evacuation plans.
We can also now predict when a hurricane will form. Once formed it can already be too late to organise proper responses. We are also looking at the relationship between weather and traffic planning. For example, 70% of traffic hold-ups are said to be weather-related.
Another partner on display was Swiss-based Siemens Building Technologies, which is already providing connected buildings, but which has a goal of using Watson to help develop digital buildings. These will not just be consumers of energy but also a managers of energy use. As Matthias Rebellius, nnnn, put it:
The pre-intelligence in the building itself will then be connected to the rest of the world and its wider issues. It will be a connected part of several interconnecting ecosystems. Our customers want both more insight about their buildings, and want to have to care less about them. That means the opportunity is there for a partner to meet both goals.
As an example of the complexity and richness of data generated by IoT, Laurent Martinez, of Airbus Industrie, pointed out that while the company's first aircraft, the A300, had some 20,000 measurable parameters its very latest offering, the A350, now has over 400,000 of them. So the company needs to be able to offer airline customers the richest capabilities in analysing the data and turning it into actions and decisions of value.
For example, he suggested that it is now time that aircraft maintenance stepped out of the 20th Century, where it is a scheduled event regardless of need.
Now it can be customised much more, depending on the data coming from sensors. We are now starting to use Watson to implementing this capability.
This is certainly one of the more important announcements in IoT. Managing the systems is an important capability, but making sense of the vast amounts of data to create actionable insights, actions and decision options could be the one factor that makes real business sense of it all.