Living on the edge
- June 11, 2026
- Steve Rogerson

Steve Rogerson visited this month’s Advantech World Partner Conference in Taipei and found everyone talking about edge AI.
Is edge AI the new IoT? That, on the face of it, may seem a silly question. One follows almost naturally from the other. As the IoT grows, more data are gathered at the edge and, for quick answers and filtering, there needs to be some intelligence dealing with those data where they are produced.
An autonomous vehicle, for example, may have a sensor detecting an object in front of the car. It can’t wait to send the images back to the cloud for them to be processed and deciphered before a decision is made. Even the thought reminds me with a smile of former Formula One world champion Lewis Hamilton last season getting used to his new team at Ferrari when he would radio race tactics suggestions and they wouldn’t get back to him for a few laps, when it was clearly too late. He could have lived with a few seconds delay, but an autonomous vehicle needs split second decisions.
There are similar applications in the likes of factories, medical and so on where the vast amounts of data being collected need to be analysed at the point of collection. So, why am I asking if edge AI is the new IoT?
That is because at the many Advantech World Partner Conferences (WPCs) I have attended over the years, I have come to realise what may appear odd at the time has actually over time become mainstream. I have mentioned before that it was at one such conference that I first heard the expression “internet of things” and thought that would never catch on. There have been other examples.
This year in Taipei, Advantech (www.advantech.com) worked its WPC around the Computex trade show, starting with an edge AI conference the day before Computex and having the full WPC at the same time as the second day of Computex. The arrangement seemed to work.
It was during the WPC when I was talking with an Advantech executive and explaining what the IMC was that he asked why we were still talking about IoT. “It is all edge AI now,” he said.
That took me back a little, I must admit, and that is when I started thinking maybe he had a point. The crucial part of IoT is connectivity. All these devices, or things, connected through the internet or whatever and acting in a coordinated way are what the IoT has always been about. If the intelligence is all taking place at the edge and the connectivity merely used to back up the data or for longer term trend analysis, then does the term IoT fit any more?
We are not at that point yet, and as far as I can see the term IoT is still very relevant. But for how long? We saw it merge with AI to become AIoT for some applications, but now the “oT” bit is often dropped.

That edge AI is a thing was summed up nicely at the conference by Ed Doran, vice president of the Edge AI Foundation (www.edgeaifoundation.org); I didn’t even know there was an Edge AI Foundation. It is based in California.
Anyway, Ed said: “With edge AI, what we are talking about is building on all the advances in the past few years and putting them in the real world. So not just the factory but the smart containers, the submarines below the waves, the vehicles moving on the ocean, the surface of the Earth, in the air, the devices in your pockets. These are all edge AI. They can warn about floods and wildfires. We are taking the intelligence we have built in data centres and bringing it out into the real world.”
He said he found edge AI “incredibly exciting” and there were greater advances coming. “Everything is changing very quickly,” he said.

Miller Chang, president of Advantech’s embedded sector, added: “The world is now at a critical inflection point for edge AI and is about to enter a major wave of rapid physical AI growth. AI has evolved from cloud-based model training and generative AI into the physical AI stage, where systems are capable of sensing, reasoning and taking action. In the future, AI will be able to participate in real-time industrial decision-making and autonomous operations, becoming a key force driving the intelligent transformation of industries.”
That edge AI is a thing is not in doubt, but is it a new term for the internet of things? Probably not. Edge AI is a crucial part of the IoT and always will be. The two terms will, as they should, live together along with all the other acronyms and jargon that are part of this exciting technical revolution.










