A center in Munich is part of a $3 billion investment to bring Watson’s cognitive computing power to the internet of things, IBM says.

IBM Corp. has been investing big in its Watson supercomputer, and a center the technology company opened last year in Munich seeks to bring Watson to the growing internet of things, says Niklaus Waser, vice president of Watson IoT Europe and the Watson IoT Center in Munich.

The center, which opened in December 2015, is part of IBM’s broader $3 billion investment to expand the reach of Watson’s machine technology. IBM is investing $200 million in the Munich center, Waser said in an interview as part of a media tour last month sponsored by Germany Trade & Invest, Germany’s economic development agency.

Research and advisory firm Gartner Inc. defines the internet of things as “the network of dedicated physical objects (things) that contain embedded technology to sense or interact with their internal state or external environment. The [internet of things] comprises an ecosystem that includes things, communication, applications and data analysis.”

IBM’s Watson uses an algorithm that learns when it’s exposed to new data instead of being explicitly programmed by an individual—a process called machine learning—which enables it to engage in a dialogue with people, learning over time how to improve its answers.

The center in Munich marks the largest IBM investment in Europe in more than 20 years, IBM says. The company is still staffing the internet of things unit and plans to hire around 1,000 developers, consultants, researchers and designers to work at it, Waser says. Today, 6,000 global clients are tapping Watson IoT services—up from 4,000 just eight months ago.


Companies including Deere & Co., the manufacturer of John Deere agricultural equipment; consumer electronics maker Panasonic Corp.; and appliance manufacturer Whirlpool Corp. are all working with the Watson Munich center for internet of things-related projects. Whirlpool, for example, uses Watson to enable its home appliances to connect and interact with one another and with consumers. For example, a Whirlpool washing machine will communicate directly with a Whirlpool dryer letting it know what kind of laundry load to expect and the optimum drying program to use, helping to reduce energy consumption. Whirlpool and Watson are teaching appliances about how people use them to aid in Whirlpool designs, and IBM and Whirlpool also are working on programs that will automatically reorder detergents, filters and other supplies directly from online retailers for consumers.

IBM says there are more than 9 billion connected devices operating globally, generating 2.5 quintillion bytes of new data daily. However, Waser says nearly 90% of that data is never used. The market for making sense of these devices is expected to reach $1.7 trillion by 2020, IBM says.

IBM offers companies four Watson internet of things application programming interfaces (APIs). They include:

Natural language processing: Enables consumers or company employees to interact with systems and devices using simple human language. For example, a technician working on a machine in an e-commerce warehouse might notice an unusual vibration. He can ask the system “What is causing that vibration?” Using natural language processing, the system will automatically link words to meaning and intent, determine the machine he is referencing, and look up recent maintenance to find the source of the vibration and then recommend an action to stop it.

Machine learning: Can be applied to any data coming from devices and sensors to automatically understand the current conditions, what’s normal and expected trends. It also can suggest actions when an issue arises. For example, machine learning could monitor data from a fleet of delivery trucks over time to learn normal and abnormal conditions.


Video and image analytics: Monitors data from video feeds and images to identify scenes and patterns. This knowledge can be combined with machine data to understand past events and emerging situations. For example, video analytics monitoring security cameras might record a forklift in a warehouse infringing on a restricted area, creating a minor alert. A week later a machine in that area may begin to malfunction. With video or image analytics, the two incidents can be connected and thus help companies better analyze and solve the problem.

Text analytics: Mines text, including transcripts from customer call centers, maintenance technician logs, blog comments and tweets to find correlations and patterns in these vast amounts of data. For example, phrases such as “these shoes are too narrow” and “these heels are not true to size” on Twitter or Facebook about the same pair of shoes can be linked, helping a retailer or manufacturer spot a product issue.

About a dozen global e-retailers are using Watson’s machine learning technology, including outdoor gear retailer The North Face, owned by VF Corp., No. 92 in the Internet Retailer 2016 Top 500, and 1-800-Flowers.com Inc. (No. 57). 166 retailers in the Internet Retailer Top 1000 use IBM technology, according to The 2017 Leading Vendors to the Top 1000.