Every week there seems to be a new revision of how many Internet of Things things there will. So to put an end to this I am now publishing my own version which is future proof: “By 2050 the number of connected devices will be huge ™”. In any case, we all realise that the Internet of Things will create a data tsunami of unimaginable size and that the opportunity lies in making sense of all that data, in finding the signals in all of that noise.
Today IBM announces the creation of a new global headquarters for Watson IoT in Munich where one thousand data scientists, researchers and programmers will work in this domain. Yes indeed, the name is Watson IoT. Watson is course of Jeopardy fame but in the mean time it (he?) has evolved into a cognitive computing solution with deep knowledge in many industries. The only way to take advantage of all the sensor data will be with a system that can learn and reason on a massive scale while presenting an easy, natural way for humans to interact with it.
We are also adding new Watson APIs onto the IoT Foundation platform. The coolest thing about these services is that they are easy to use while under the covers they are highly complex. No need to be a data scientist or mathematician, just send the request and take the response. I don’t have the details yet but the new services will allow you to build applications that
- interact with the system in human language. Using natural language processing to help understand the intent of human language by linking it to other sources of data to put it into context in specific situations.
- look at data coming from devices, sensors and other sources to understand what is ‘normal’, what is a deviation and what is the trend in order to propose actions when necessary.
- find correlations and anomalies in video feeds and image snapshots to identify scenes and patterns.
- discover potential issues in a certain domains by reading maintenance logs, blog comments, etc. using text analytics.
Here is the video of the announcement.