A whiff of the future at CES 2024

A whiff of the future at CES 2024

Although most of the attention at Las Vegas’ annual Consumer Electronics Show focuses on product releases from the big companies – LG, for example, debuted an OLED television fabricated on a completely transparent substrate, so you can peer through it like a window – the most interesting and innovative takeaways can reliably be found amongst the thousands of startups and tiny companies vying for the attentions of the 200,000+ CES attendees. After 5 days, countless demos, and thousands of booths, here’s some of what really impressed me at this year’s show:

Electronic chip on blue surface
Tetraram’s memristor. Credit: Mark Pesce

Tetraram’s Memristor: The ‘memory resistor’ (or ‘memristor’) is a basic electronic element that has the capacity to ‘set and remember’ its resistance value. Unlike passive resistors, a memristor’s resistance can be modified, in the presence of a magnetic field. First described in 1971, forty years passed before working memristors emerged from the lab – into a world of digital ones and zeros, broadly incompatible with the ‘in-between’ values offered by memristors.

At least, that was true until the modern age of artificial intelligence: most applications of artificial intelligence are built atop artificial ‘neural networks’ whose internal connections have ‘weights’ – values that lie somewhere between zero and one. Having a way to store these very non-digital values in a circuit element – such as a memristor – brings enormous efficiencies to energy-hungry AI applications.

Tetraram demoed a first-generation AI ‘accelerator’ using their memristor technology, and reckon their chip makes AI at least 10x more energy efficient – a welcome improvement as AI works its way into everything. (Here’s their Nature paper describing their tech.)

Models of brains, people and eeg chip on table
WIMagine. Credit: Mark Pesce

WIMagine: 2023 saw a lot of advances in the sorts of bioelectronic interfaces that enable paraplegics to walk. The University of Grenoble has led the way in a lot of this work (here’s their Nature paper), and I got a close look at their brain interface. While it requires surgery to implant into the skull, their sensor functions more as a permanent set of EEG electrodes reading the patient’s brain.

That continuous EEG provides a signal path between the patient and the bioelectrical systems that allow the individual to move, speak, and so forth. The whole procedure leaves the nervous system intact, the brain unmolested within its protective shielding – and it also means that the brain’s natural defences don’t slowly coat any intrusive electrodes, diminishing their effectiveness over time.

That means the system should be a ‘one and done’ operation for the patient, who then spends months training themselves to ‘talk’ through the EEG interface, and through that interface, to their bodies. This looks like a sustainable approach to bringing a lot of capability to people who would otherwise be trapped in their own bodies.

Person at convention centre wearing black armband with electronic devices attached holds hand out to group of people
Gyrogear’s hand stabiliser. Credit: Mark Pesce

GyroGear Hand Stabiliser: As the world’s population ages, we’re seeing a lot of innovation coming out of ‘Agetech’ – technology to help support us as we grow older. One symptom of aging is motor tremors. Common in both Parkinson’s and Parkinsonoid conditions, these motor tremors make it very difficult for people to perform manual tasks, and have a significant impact on quality of life.

A research group out of Singapore – GyroGear – has come up with an unexpected solution: a glove containing a sophisticated system of sensors and gyroscopic counter-balancing that effectively ‘cancels out’ most of the arm’s muscle tremors. Think of it as noise cancellation, but for your neuromuscular system. If it lives up to its promise, GyroGear is going to restore quality of life to a lot of folks. (Honorable mention to NextStride, an almost impossibly elegant solution to helping people with Parkinson’s walk confidently – by giving them a guide light!)

Drone mounted on wall
The Jedsy glider. Credit: Drew Smith

Jedsy Glider: Five years ago, on my last visit to CES, almost one entire hall at the Las Vegas Convention Centre had been filled with drones and drone-like products. That hype has died down, leaving some clear air for real innovations. Although ‘quadcopter’ drones are now being tested for short-haul deliveries, they generally don’t carry enough on-board power for flights beyond a few kilometres.

Swiss firm Jedsy rethought the drone, blending it with aerodynamic elements copied from gliders, resulting in an autonomous aircraft that can travel for as much as 100km – much, much further than any quadcopter, and making it ideal for the rapid transport of medical samples and other small but high-priority items. Capable of transporting up to 3kg, Jedsy’s hybrid glider drone looks like the future.

Shiny box with pipes attached to glass vessel
Sharp’s olfactory sensor. Credit: Mark Pesce

Sharp’s AI olfactory sensor: Finally, my biggest surprise at CES came from Japanese electronics manufacturer Sharp. Out of their labs in Nara, Japan, has come a stunning innovation – an olfactory sensor. Although we’ve had microphones for nearly 150 years and video cameras for more than half a century, we’ve never had any sort of sensor that can capture our most foundational and deeply wired sense, smell. We still don’t have a complete understanding of our own olfactory system, and attempts to produce olfactory sensors have never worked very well.

Sharp ‘brute forced’ its solution, fabricating a 10cm x 8cm device that operates as a miniaturised spectrograph. The device ‘breathes in’ a sample, then runs high-voltage electrical current through it, creating a ‘map’ of the molecular components with the sample. That map – the ‘smell’, digitised – is then referenced against a massive data set of smells that Sharp’s system has been trained to detect, using AI inferencing to produce the nearest and therefore most likely match. Right now, Sharp plans to use it to help oenophiles measure the bouquet of their wines, but if and when it becomes smaller and cheaper, this could find broad use in a wide range of both medical and consumer settings. It’s a new sense – a digitised nose – and exactly the sort of innovation I hope to find at CES.

That’s all for this year. Time to hang up my conference badge and ponder all the innovations we might expect to see this time next year. Will there be more AI? Almost surely. More Agetech? Yes, absolutely. And I’m sure that I’ll see something unexpected and surprising – CES manages to do that, every time I attend.

Please login to favourite this article.