November 14, 2014 | Written by: Nathan Vega
Share this post:
Estimates for the number of connected devices in 2020 range from 26 billion to 30 billion, but who’s splitting hairs on this? It’s a lot of devices and it’s going to mean the vast majority of things we interact with will have a unique Internet identity. To keep things very simple in this blog, think of IoT devices in two broad categories. There are the data capture devices and there are devices that interact with the environment. As an example, a heat sensor captures the temp and transmits it whereas a Nest thermostat has an ability to capture AND alter the temperature. Our upcoming webinar on November 19th at 11 am, Using Watson to Build Cognitive Internet of Things Apps on Bluemix, will introduce the concepts of IoT and we’ll show you how you can use Watson to interact with your IoT devices.
This is a seminal point for future smart devices that are not simply “smart” because they’re connected, but smart because they can react and even anticipate our needs.
Why all the fuss about Internet of Things?
The first networked device was a Coke machine modified by students at Carnegie Mellon University in 1982. The story goes that the students wanted to see if cold soda was available before they made the walk. That’s 32 loooooong technology years ago. Yet, here we stand in 2014 talking about a burgeoning group of “smart” (aka networked) devices. And, many times I have the same reaction I had to the first story..Sounds kinda cool, but it’s more novel than practical. What the students stumbled upon is a problem every enterprise and individual faces daily. We need accurate, voluminous, and reliable data for decision making. The problem is that, by and large, the information on the Internet today is generated by humans. The images, the blogs, the tweets, wikis, etc are generated by a notoriously flawed source, humans. We have challenges with accuracy, time, and representation of data.
So, creating devices that can capture this information has served a great purpose in solving the challenge of volume, accuracy, and reliability. Yet, this is just the cusp and now we have a big data problem.
We have a lot of accurate data, now what?
So, we journey forward a couple of decades and the practical element of IoT which is data capture has been used for years in industry to keep machines humming, inventory tracked, and operations inundated with more data than they can interpret. We’ve built systems that interpret the data points and do some reactionary things like notify an operator, blink a light, shut something down, etc. These are a sophisticated set of systems designed to interpret data streams and to do something with them, but they still are missing something. They lack a deeper understanding or context. They simply know that if X happens then do Y. There is a great opportunity to evolve this with complimentary systems like Watson.
Cognitive Systems will mediate the future
A cognitive system like Watson interprets underlying meanings, understands concepts, and hypothesizes a response. This means it stands in a great position between us and the world of connected devices. Imagine a machine that understands your Facebook post that you’re seeing auras, having reactions to light, and noises as a migraine. That same system could do more than just notify you of a migraine. It could adjust the environment in your house to lower lighting, mute everything, and run your fan to get rid of fragrances before you get home. Imagine your rental car radio pre-programmed to the local stations you’d like based on your current mood. These concepts are more fact than fiction and we want to show you the Watson and Internet of Things services available on Bluemix are your place to get started.
Please join us on for our webinar, Using Watson to Build Cognitive Internet of Things Apps on Bluemix, where our speakers will show you how you can use Watson to control IoT devices. We’ll have a live, working demo with code available for you start playing with. What will you do with Watson?