Artificial Intelligence and the Internet of Things

January 16, 2024 | Jay Slade, Vice-President, Analytics and Business Intelligence, RBC Dominion Securities


Hands typing on laptop.

In my last blog, “AI: Three reasons why there’s so much hype”, I explained how AI is, at its core, just math. As such, AI is truly a tool for many different use cases – and humans (not machines) put their own values and ethical judgements into those use cases. The reason why AI has been getting so much attention lately is because of the recent advancements in computing technology and the availability of massive amounts of data for AI model training.

So how is AI connected with the “Internet of Things”? This is the concept of devices used in our everyday lives being connected on a network of some kind, like your home wifi or cellphone. Think for example about home security systems, smart speakers, thermostats, or even cars. Your device can send information to you on another device you have with you, and you can send instructions back, like increase temperature of the house, or lock the door. Devices are even starting to talk to other devices, e.g. your refrigerator tells your smart phone that you are running out of a certain food.

So what role will AI play in this growing ecosystem of connected devices? Imagine a brand-new subdivision being built from the ground up by a single housing developer. However, this developer not only builds houses but also puts in common smart appliances and devices into all the homes. So every appliance in the house, the security systems, lighting, heat, water usage – everything – is a smart device that receives and transmits data.

Now imagine that each house anonymously sends usage data to a central server that the developer has access to. With that type and volume of information, an AI model could learn, for example, how to create the most energy-efficient household profiles for the subdivision. It could understand when to power things on and off, when to turn lights on and off, etc. The server could communicate with each household, sending daily recommendations as to how to manage that household. This may sound intrusive to some but consider the busy families that are bolting out of the house to get to work or attend children’s activities, forgetting to lock the door. Or a person with cognitive issues forgetting to turn off the stove. In fact, different AI models could be used to suit different family situations.

You could think of AI as a backup system. You run your house the way you want but the AI can take over when you are on vacation, etc. The possibilities and use cases here are endless, but it all starts with the concept of devices transmitting data and the usage of AI to learn from that data and produce useful instructions to give back to those devices.

A final note on this topic. In my household there are a lot of lights to turn off every night, which takes us about eight minutes. Sure, we could make our lives a bit easier by having wifi-connected power outlets connected to our phone to shut things down. But there are times when I think it would be nice to just let the AI figure out that midnight is our usual bedtime and start shutting things down for us. The point is, AI is a tool that can “learn” based on your needs – and it has use cases that can extend beyond your computer to the things you use in everyday life.