Jake Porway came back to talk about the coming explosion of data. With the Internet of Things, drones and more sensors everywhere, our current view of “big data” is going to seem quaint in the not too distant future. So much more data is going to come flooding in we have to adopt new technologies and new analytic approaches.
With that, Beth Smith came on to talk about the Internet of Things with a story about prosthetics. Her husband uses one and in the past he had to adapt to the way it worked. Today his new one has a computer built into it allowing it to work the way he needs it to – the prosthetic adapts to him rather than the other way around. Medical devices that collect data in real-time, analyze this data and feed it back to improve the outcomes for patients are increasingly common she says. This kind of pervasive instrumentation is what is driving the Internet of Things.
For some examples, Beth talked about some IBM customers. For instance Peugeot has connected its cars to provide new web-based services for drivers on everything from finding gas to car maintenance. A Chinese power company using sensors to improve its forecasting of renewable power generation based on the weather data it is collecting. The trick, she says, is to turn the internet of things into insight. Three simple steps she says
- Capture data streaming from large numbers of devices.
- Integrate and rapidly analyze this data in combination with all other kinds of enterprise data.
- Act on this at scale.
To deliver on this promise IBM has delivered a set of services on BlueMix – an IoT foundation, the Cloudant JSON database, time series services and more. Mike Curry joined her on stage to do a demo of the IoT services on BlueMix. He walked through how the IoT foundation and Node Red lets him link a mobile phone so it can be issue a proximity alert if you go somewhere you should not. Lots of good stuff in terms of easy connectivity, data capture and analysis. Sadly he wrote the business rules for managing a critical decision – is this person doing something unsafe – in the mobile app instead of managing them in a proper decision management backbone. Poor design approach. What would work better would be to have the mobile app call a decision service that would have determined which of the appropriate actions make sense.
A customer story followed, discussing how sensor data is being collected to analyze trends and behavior in retail. Not clear if the data was being acted on in real-time but it’s clearly being used to drive analytics.
Doug Balog came on to wrap things up with some discussion of infrastructure and how important having the right kind of infrastructure is going to be moving forward. With more data, more of it streaming, companies need new infrastructure capabilities.