Fullscreen HTML5Med Res (??MB)HTML4

Techno Talk Scan me! From Dr McCormick’s mercury thermometer to AI-powered diagnostics, I take a look at how yesterday’s science fiction is quietly becoming today’s reality I’m constantly amazed at the pace of medical progress. When I was about five years old, in the early 1960s, I had some trouble breathing. My mum took me to our family doctor, Dr Mc­ Cormick. He said that I needed to have my tonsils removed. He also moaned a bit about the fact that the rules had changed and he would have to refer me to the hospital. In the good old days, he said, he would have performed the operation himself in his surgery. I can only say that I’m glad I wasn’t born in the good old days! All I can remember about his surgery was a height-measuring rod attached to a weighing scale, a blood pressure cuff, and a wall chart showing parts of the human body (in case he forgot where they were). The only other medical equipment I remember seeing was the stethoscope hanging around his neck, the wooden tongue depressors he used to make me gag, and a big mercury thermometer he stuck in my mouth while instructing me not to talk and then asking me questions. Then there’s now Anything else in the early 1960s required a referral to a specialist or the hospital. By comparison, during a visit to my doctor’s surgery several years ago, I noticed the outer door to their reception area was making an annoying squeak, so I returned the next day with a can of WD-40 (I guess WD-1 through WD-39 didn’t quite do the trick), which addressed the problem. That squeak must’ve been annoying the receptionist too. Ever since then, if I ask, “Is there any chance you can squeeze me in”, I usually manage to get an appointment within a couple of hours. By some strange quirk of fate, this happened just yesterday. For the past couple of days, I’ve had a cough that’s been making my eyes water and my chest ache. The pain in my chest was so intense I started to worry that I was having a heart attack! Happily(?), it turns out that my coughing was so bad that I’d pulled a muscle. The reason I’m waffling on about this is that the doctor (well, one of her minions) gave me a chest X-ray and an EKG. They also took a nasal swab to rule out the flu and COVID, and they drew 74 some blood and ran a batch of tests in real time in their on-site lab. Although they didn’t use it in this case, they also have an ultrasound machine and goodness knows what else in the surgery. All of which makes me wonder what doctors in the early 1960s would have thought if they could see a modern family doctor’s surgery. I suspect they would have assumed they’d inadvert­ently wandered onto the set of a sci-fi film. The knee bone’s connected to… When my dear old dad had a hip replacement operation in the 1970s, he was confined (physically restrained) to his hospital bed for weeks before they let him start walking. By comparison, one of my friends who is in his mid-70s had a hip replacement just a couple of months ago. He went into the hospital early one Tuesday morning. I called him in the evening to see how he was doing. He was already back home. I’d caught him climbing the stairs to his bedroom! Starting about ten years ago, several of my contemporaries began to need knee replacement operations. My eyes are watering just thinking about this. They all said that the first couple of weeks after the operation were extremely painful, but after a couple of months, they felt like they’d been given a new lease on life. Once again, by comparison, two more of my friends have had knee replacement operations recently. Apparently, there is a new nerve-blocker technique, because both were walking the day after the operation using nothing more than over-the-counter pain medication. Scan me! Many of the technologies that were the stuff of science fiction when I was a kid don’t even raise a quizzical eyebrow today. This makes me wonder what we might expect in the not-toodistant future. Dare we hope that one day we’ll each have handy-dandy little AI-powered diagnostic machines in our homes? I’m envisaging being able to insert my finger to have a drop of blood taken and receive a diagnosis of any common ailments just a few seconds or minutes later. I’m also looking forward to the day Max the Magnificent when my doctor has a full-body scanner in her office. I imagine stepping into a booth, standing still for a few seconds, and then watching a detailed 3D model of my innards appear on a screen, all displayed in glorious colour. Of course, the real magic wouldn’t be the pictures themselves. It would be the AI systems working behind the scenes, comparing my scan with millions of others, spotting tiny abnormalities that no human physician could reasonably be expected to notice and highlighting potential problems long before they started causing symptoms. The best part would be catching things early. Instead of waiting until something hurts, breaks, leaks, swells, or otherwise announces its presence in an unpleasant fashion, doctors could identify many problems while they were still small, simple and easy to treat. Futuristic technologies Traditional industrial robots are astonishingly precise, but they are also remarkably clueless. Put a car door in exactly the right place, and a robot can weld it all day long. Move the door a couple of inches or centimetres, however, and the robot may suddenly find itself having an existential crisis. One company working to change this is CynLr (www.cynlr.com). Its goal is to give robots something approaching human-like object intelligence. Instead of relying solely on pre-programmed instructions, these systems learn by interacting with objects, observing the results of their actions and gradually building an understanding of how things behave. Imagine a robot encountering an unfamiliar object. Rather than simply seeing a collection of pixels, it begins to recognise concepts such as shape, Practical Electronics | August | 2026 weight, balance, movement and purpose. It learns that a cup can be picked up by the handle, a drawer can be opened by pulling, and that some objects are fragile while others are not. This may sound like a small step, but it represents a profound shift. Robots that can understand the world through experience could eventually work safely alongside humans, adapt to unfamiliar situations and perform tasks that are currently beyond the reach of conventional automation. As the folks at CynLr said, “We are teaching machines not just to see the world, but also to understand it.” Where are my keys? Speaking of things that would have sounded like science fiction not so very long ago, consider AI-powered glasses that remember where you left your keys. One company exploring this idea is Brilliant Labs (https://brilliant.xyz), with its Halo smart glasses and Noa AI assistant. Halo looks remarkably like an ordinary pair of spectacles. However, hidden inside are cameras, microphones, bone-conduction speakers, a tiny display and enough computing power to make me squeal in delight. According to Bobak Tavangar, the company’s CEO, the system can see what you see, hear what you hear, and engage in real-time conversations about the world around you. What really caught my attention is Noa’s memory capability. Rather than simply answering questions, it builds a personalised knowledge base based on the wearer’s experiences. This means you could ask questions like, “Where did I leave my keys?” or “What was the name of the bald guy I met at last week’s conference?” and have a reasonable expectation of receiving a useful answer. As Bobak told me, the goal is not merely to create smarter glasses, but to create a kind of memory assistant that helps you keep track of your life. As someone who spends an inordinate amount of time at my workbench looking for tools and components that ‘were here just a minute ago’, I can definitely see the appeal! Itera’s liquid circuit board concept. through which liquid metal can flow. By controlling where the metal moves, the system can create, remove and reroute electrical connections on demand. The result is a circuit board that can effectively rewire itself. Need to connect two signals? Create a path. Need to disconnect them? Remove the path. Need to route a signal somewhere completely different? Reconfigure the board. This may sound like Terminator-2level technology, but Itera has already demonstrated working hardware and attracted significant interest from industries ranging from automotive and defence to hyperscale computing and semiconductor development. As Itera’s CEO AJ Cooper explained to me, the long-term vision is to make hardware development dramatically faster and more flexible. If successful, engineers could one day modify physical circuitry almost as easily as updating software. As someone who has spent countless hours redesigning and respinning circuit boards, I find this to be a remarkably appealing prospect! A blast from the past Sometimes, we become so fixated on what is to come that we inadvertently overlook what has already been. While most companies are focusing on ever-larger artificial neural networks (ANNs), the answer to running AI on low-power edge devices may lie in a remarkably different approach with roots tracing back to research conducted in the Soviet Union in the 1960s. One company championing this idea is Literal Labs (www.literal-labs.ai), whose technology is based on Tsetlin Machines. Unlike neural networks, which learn by adjusting vast numbers of numerical weights, Tsetlin Machines learn using collections of simple logical rules. The result can be AI models that are dramatically smaller, faster and more energy-efficient than their neural-network counterparts. This is particularly important at ‘the edge’, where devices often have only a few kilobytes of memory and must run for years on tiny batteries. In demonstrations presented by Literal Labs, Tsetlin Machine implementations have achieved comparable accuracy to neural network solutions while consuming a fraction of the memory, processing power and energy. While much of the AI world is racing toward ever-larger models containing billions or trillions of parameters, the future of edge AI may emerge from an elegant idea that has been hiding in plain sight for more than 50 years. PE A circuit board that rewires itself Once traditional PCBs (printed circuit boards) have been designed and manufactured, their electrical connections are fixed forever (well, unless you wield a scalpel, a reel of wire and a soldering iron). If you want to make more than minor modifications to the circuit, you usually need to spin a new board. Itera (www.itera.co) is working on a radically different approach. Its technology is based on multiple layers of ultra-thin glass containing microscopic channels Practical Electronics | August | 2026 75