The IoT is All about AI
In the ongoing quest to connect sensors, devices, and systems, it’s easy to overlook just how profoundly artificial intelligence (AI) and machine learning intersect with the Internet of Things. Today, the data that streams in from smart lighting, HVAC, weather stations, cars, smartphones, RFID chips in factories, and more fuels the digital revolution.
All major technology vendors, including Amazon Web Services, Google, Oracle, Microsoft, SAP, Salesforce, and GE, have stepped aboard the AI-IoT train. They are building an IoT-supported framework into apps, analytics tools, and enterprise applications. These AI integrations range from image and voice processing through sensors to AI on the network edge and in the cloud. The common denominator is that the right set of data points arranged the right way can produce transformative results.
Gartner has referred to this emerging environment as a “digital mesh” that will ultimately embed AI in virtually every IoT device, blending the digital and physical worlds. This framework is all about exploiting connections between expanding sets of people, devices, content, and services, the analyst notes.
On the Edge
By now, it’s fairly apparent that AI and machine learning can transform data into information and, ultimately, knowledge. It helps make dumb machines smarter, automates processes, and adds remarkable value to systems. As organizations put algorithms to work, tackling things as diverse as how to optimize fleets of aircraft or trucks or when to perform maintenance on a machine, they achieve real-world cost and efficiency gains.
But, increasingly, there’s a need to react to conditions faster and manage complex environments, such as a smart transportation grid, more effectively. This is prompting organizations to look at embedding AI at the edge, where it’s possible to reduce latency between the moment an event occurs and when action follows. Altering this control loop can shave precious milliseconds off reaction time. It enables entirely new processes and possibilities.
AI, edge processing, and fog computing used with the IoT also enable more specialized apps and plugins that accomplish specific tasks. Some examples might be ensuring integrity within an IT or IoT framework, including links between sensors and controllers; and managing these links when a particular device, link, or communications system fails. AI has the power to detect problems and deliver fixes with little or no human intervention. It also has the power to sense changes and anticipate failures and cybersecurity risks.
AI + IoT = $$$
All of this means systems integrators, managed services providers, and others must consider the role of AI and machine learning in every project, now and in the future. This entails building robust, resilient, and flexible IT and IoT networks, as well as gaining a thorough understanding of how to deploy and use sensors, how to design new network topographies (such as edge and fog), and how to take advantage of emerging technologies such as 5G.
An AI-centric IoT framework represents a next step in the evolution of digital technologies. It will fuel massive changes across areas as diverse as agriculture, manufacturing, retail, healthcare, law, transportation, and financial services. System integrators that gain expertise in the field and understand how to design and build leading-edge AI-IoT frameworks will unlock value and gain a competitive advantage.
Samuel Greengard is a business and technology writer based in West Linn, Ore. He is the author of The Internet of Things (MIT Press, 2015).