TRENDS AFFECTING AI
Gartner expects that three trends will affect AI in the next few years. One, better communication (both ways) with people: Natural-language processing, generation, and contextual interpretation will make AI more comfortable to use and will improve the use of all computing resources.
Secondly, more in-depth and broader integration with existing applications and IoT projects: AI has its most significant value when built into architectures that drive business and service value.
Third, richer ecosystem interaction. As AI becomes more common, applications that employ it must work effectively with others using similar technologies, which will result in chains and meshes of AI systems that work simultaneously toward their individual goals in a cooperative but decoupled fashion. Generally, AI will trend from one-off experimental projects to an approach that integrates the technology with the business
The era of the 90’s started with something called the machine to machine or the M2M solutions. Back then, there was a lot of overlap in the processes involved before telecom and wireless helped facilitate a lot of segregated stuff.
Still, there’s lots of overlap with mobile device management and IoT platforms, small devices, getting some help taking it back to a server doing something about it sending apps and commands down to the device.
Coming from the machines having instrumented machines, you have a lot of people saying the same things. I want predictive failure analytics. I want to see what the remaining useful life on this machine is. I don’t want to go by the user manual to tell me when I need to replace parts or things like that.
ML, AI & SENSORS
People are trying to save money, as usual, extend the life of things. So, machine learning has two levels – level one and level two AI. Level one AI is what people commonly think of as machine learning. You’ve got people doing creating scripts, or Python, our Weka, or things like that.
Businesses typically have a data scientist paired with a domain expert for the machine. They pick the right algorithm, get the right model, figure out all the moving parts, and then train the model. And that is the scoring engine. And they’re hoping that they’re going to get the right answers to tell them that a particular piece of equipment will fail next Thursday, and maybe you should fix it before it fails.
So, all of those pieces of equipment are covered with sensors. And that sensor data is what’s feeding AI? So, if AI can bring inputs, run it through to make decisions based on all of the inputs, look at big data analytics, decision trees, and then action can be taken.
AI IS THE CRUCIAL PART NOT IoT
The IoT part is the easy part though getting things connected is still a hassle. But that’s the lowest value part of it, no doubt about it, the analytics, the AI is the most crucial part. What’s cool is this level two AI, which you don’t have to have a domain expert. One can say that this is the kind of outcome I’m looking for and feed the data, and it figures it out on its own. I think that’s going to be a game-changer. And a big differentiator for anyone.
A lot of the stuff which is now cognitive and is the center for the future of work is looking at how all these new developments in technology and innovations impact strategy?
So we’re looking at the Internet of Things, artificial intelligence, and across all the industries saying, how does that now impact strategy? It’s easier to do Software as a Service. Still, hardware as a service is a more challenging thing.
Whenever you hear someone show up and say, this will transform your business that sounds ambiguous. But with instruments and equipment, and great AI, and a great back platform, you actually can’t and transform your business. Any day, service based on outcomes based on predictive analytics will help one stay ahead of the curve. It’s the way of the future, and it’s right to the point. Transformative.