Wireless Technologies

IBM's big bet on 'cognitive computing enabled' IoT

In slew of high-profile partnership announcements, IBM officially opened its new $200M global headquarters for its Watson Internet of Things (IoT) business in Germany, yesterday. Housed in a high-tech Munich tower, the aim is to facilitate real-world, industry specific, enterprise IoT solutions which incorporate cognitive computing.

The strategy, explained John Kelly, SVP of cognitive solutions at IBM Research at the accompanying event, is to offer a cloud platform, lead in AI and do it through an industry lens. “Cognitive Watson is one of the biggest bets IBM is placing,” he said.

This does seem to be a very smart move for IBM. This is because firstly, both AI and IoT are probably the hottest trends around. Secondly, they are each areas that cause some confusion around how they might really work, and here, IBM is looking to provide a clear roadmap to secure, usable enterprise solutions.

In practice, IoT is difficult to actualise because it involves both hardware and software. The feedback received from most of the industry professionals I’ve spoken to over the last year or so is that success involves a lot of collaboration. This is precisely the approach that IBM has opted for. And of course, IBM is big enough to attract some very strong partners. 

As ever though, most of the use cases discussed covered the standard B2B examples. Sensors on aeroplanes, enabled train tracks and smart street lighting all, of course, got a mention. There was the first cognitive computing enabled drone – courtesy of Aerialtronics – which boasts potential uses in crowd safety, damage assessment and aviation inspection. However, probably the two most interesting partnerships announced yesterday were the one with Visa and the one with crowdsourcing platform, Indiegogo.

The Visa partnership aims to help a variety of connected devices to deliver secure payments. The most talked about example was where a car becomes a point-of-sale device and potentially recommends a petrol station, takes you to the correct pump then pays for your unleaded. In practice, this is only one step up from Apple Pay but none-the-less it is a clear move towards the kind of financial changes which have been touted for a long time. It also is yet another step in the direction of new – more service orientated – revenue models.

The Indiegogo partnership, on the other hand, is a decent way of getting in on the ground floor of the next generation of innovative companies. In practice, this provides “qualified entrepreneurs” on the site with free access to the IBM Watson IoT platform for an unlimited time period. In the same vein, IBM has also been bringing in a lot of very young people in to make the physical space more “millennial friendly”. Interestingly, the girl I spoke with was extremely enthusiastic and had no idea the company traditionally had such a dry image.

Yet the real stumbling block for IBM could still be a perception one. Last year I spoke to a number of people about AI-as-a-service in the light of Infosys’ launch of its Mana project and what emerged most strongly was what a poor reputation IBM had in the AI space compared to Google. Many dismissed Watson as mere machine learning.

During his initial keynote Kelly was very keen to highlight the difference between machine learning, AI and cognitive computing [a PDF white paper he produced on the subject can be found here]. Yet whichever way you look at it, nuanced definitions never stick as well as an easy blanket perception.

When I put this to the IBM research team during a group interview Dinesh C Verma, IBM Fellow of Distributed Cognitive Systems conceded the work Watson does is “not as visible” as the work of Google DeepMind, but then again, it is not a consumer solution. This is aimed squarely at the enterprise, he explained, and the mix of abilities that IBM offers securely via its cloud platform are “more important” to clients.

So, what might clients think? Well, the launch event also included the true client experience via the standard building tour. This incorporated four different floors of the 30-odd story structure and highlighted some of the incredible expertise available.

One level – which was led by the brilliant John Cohn – was dedicated to modelling the building itself as a showcase.  Another featured the “IoT consultancy” which boasted lots of post-it notes and used design principles to quickly help deliver viable working prototypes. While in the Industry Lab, Heike Kammerer, who heads it up, explained “this is not a research lab or a showroom, although it might look like it”. Instead it is a home for real working projects – like a robotic arm that can follow voice commands.

The Watson Internet of Things new headquarters certainly provides a range of highly impressive ideas which fit the zeitgeist and, by no means, mirror the traditional picture of IBM or even a fun established R&D centre like Hursley in the UK. Every trendy feature is in evidence. Yet despite the foosball machine and fridges of soft drinks it certainly doesn’t smack of some young startup.

The one thing that did strike me though, was that for all the talk about creating this model sensor-driven environment, the building itself was baking hot. The bright February sun streamed relentlessly through the greenhouse windows and so, by the time I stepped out of the high-speed glass elevator and emerged into the fresh air, I wondered if some of those clever calculations had gone awry.


« Quotes of the week: "An appalling violation of freedom of expression."


More robots, more freelancers, more openness: Samsung's vision for the future »

Recommended for You

Trump hits partial pause on Huawei ban, but 5G concerns persist

Phil Muncaster reports on China and beyond

FinancialForce profits from PSA investment

Martin Veitch's inside track on today’s tech trends

Future-proofing the Middle East

Keri Allan looks at the latest trends and technologies


Do you think your smartphone is making you a workaholic?