Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage News IBM Brings Watson to IoT

IBM Brings Watson to IoT

IBM has inaugurated the IoT Global Headquarters and will use the Watson technology to analyze and interpret IoT data.

The Internet of Things is getting bigger every day. There are currently over 9 billion connected devices generating 2.5 x 108 bytes daily, according to IBM. The market opportunity is to reach $1.7 trillion and 50B devices by 2020, the latest being an estimation from Texas Instruments. It is no wonder big companies like Microsoft, Google, Apple, Intel, Pivotal, and others, want to make the most of this market. With a planned investment of over $3 billion, IBM wants to have an important market share in IoT.

Besides the existing eight IoT centers and 48 data centers around the world, IBM has just inaugurated the IoT Global Headquarters in Munich, Germany, a center where 1000 researchers, designers, developers and consultants are to find solutions for the world’s IoT problems. While there is an enormous amount of IoT data, 90% of it is not used, according to Harriet Green, GM of Watson IoT and Education, the new IoT division within IBM. To interpret, understand and act upon this data IBM wants to use cognitive analytics already built in Watson technologies. IBM offers four API services for the IoT:

  • Machine Learning – monitoring data, including real-time one, to discover patterns and deviations from them useful to understand if monitored devices are functioning properly

  • Natural Language Processing (NLP) – enabling systems to understand human language and act upon it, controlling various devices

  • Text Analytics – mining through unstructured text – logs, transcripts, comments, etc. – in order to find out correlations and patterns leading to the discovery of various issues in the functioning of devices

  • Video and Image Analytics – same as text analytics but for images

The IBM IoT Foundation Platform provides cloud services for developers to connect various IoT devices to applications in order to monitor, analyze and control these devices. Currently the platform supports smartphones, TI Sensor Tags, Raspberry Pi, ARM mbed, Intel Edison and Galileo, and many other devices.

According to Green, the first phase of IoT – collecting information from devices – is over. We are entering a new phase in IoT development which creates “vast communities of devices that share information, which in turn can be interpreted in a larger context and managed by people using cognitive systems. In the era of Cognitive IoT, no machine is an island.”

Rate this Article