Skip to main content

Tech Toys for Grownups! (Because we like toys, too.)

In my last post, I looked at tech – or tech-adjacent – gifts for the kiddos. Now let’s get into the holiday gifts for the grownups in our lives. Maybe even gifts for ourselves. (Just don’t tell anyone!)

As with tech gifts for children, there’s no lack of sources for suggestions for tech gifts for adults. Two that I liked were the gadget lists from Consumer Expert Tech and from The New York Times Wirecutter. Here’s the items that I found most intriguing and/or practical.

If you’re a traveler, you might want to take along a Muama Enence with you. This is a translation device that claims to turn you into a fluent speaker of thirty-six languages. Looks very interesting, but I will note that there are other translators that support a lot more languages (and get more stars on Amazon, too).

Tired of your phone battery draining at just the wrong time, but not interested in carrying around a bulky charger? The Charge Card Wireless is the size of a credit card, so you can slip it in your wallet and power-up when needed.

If you live in a place where the winters can get mighty cold – a place, like, say Syracuse NY – the VolteX Unisex Heated Vest can help keep you warm when you’re out there brushing off your car, shoveling your front walk, taking the dog out when the temperature has plummeted to single digits. (Not that I’d know anything about cold weather…)

Admittedly, the WoodRanger Mini Chainsaw caught my eye because it, well, looks like a toy chainsaw. It uses the latest in quenching technology to keep it from heating up, and it lets you take care of mini gardening and landscaping projects, one-handed.

Even if you’re all tuckered out by shoveling in your heated vest or pruning with your mini chainsaw, you still may have a hard time falling asleep. Yet who wants to take a sleeping pill unless they absolutely have to? The Chill Pill is a little handheld device that sends out electronic pulses that quickly lull you into sleep. ZZZZZZ….

I’m a lifelong engineer so you can trust me when I tell you that, with the possible exception of the mini chainsaw, I’m usually not drawn to anything cutesy. But I did get a kick out of the Victrola Rock Speaker Connect, a Bluetooth speaker that’s, well, shaped like a rock. It’s designed for outdoor listening, so it’ll fit right in.

And if the noise from the speaker – or from the mini chainsaw – gets to be a bit too much, Loop Switch ear plugs provide three noise reduction levels. Perfect for concerts, where the noise can definitely get out of control.

It won’t fit in your wallet, but the Belkin MagSafe 3-in-1 Wireless Charging Stand will let you power up three devices at the same time. (And, I hate to say it, but it looks cute.)

Bird feeder in your backyard? Probably not as techy as the Netvue Birdy Feeder, which lets you feed the birds and take photos and videos of them, and any squirrels who may chance by. And it comes with color night vision, so you can also capture any raiders – avian and non-avian alike – who visit after dark.

Sure, we had a drone on our tech toys for kids list, but what’s a list of tech toys for grownups without a drone on it? The DJI Mini 3 gets props for ease of use. The one downside seems to be that it doesn’t come equipped with obstacle-avoidance sensors like more expensive drones have. You’ll just have to pay a bit more attention to where it’s flying.

My favorite item just might be the Bambu Lab A1 Mini, a 3D printer billed as a “thing maker.” At less than $500, this is a home version that’s relatively affordable. It won’t let you print “trickier materials” – the article mentions ABS and carbon fiber – but you could still make an awful lot of things with it. And what engineer doesn’t like to make things?

That’s it for now. Wishing you happy holiday shopping, even if you’re just shopping for gadgets for yourself!

It’s that time of year: tech toys for kids

It’s that time of year when folks start looking in earnest for holiday gifts for kids in their lives. And for a lot of folks, that means toys with a technology or science-y slant. There’s certainly no lack of ideas out there, but I thought the lists from Kris Naudus on Engadget and Max Dunn on Top Tech were both pretty good. Here are the highlights.

For the really little ones, there’s the Fisher Price Laugh & Learn Game Controller that looks like a real game controller, if real game controllers looked like Fisher Price toys. This one lets those kiddos play along while you’re gaming, and learn numbers, letters, and shapes while they’re at it.

More old school than high tech, Fisher Price also has the Fisher Price Rockin’ Record Player. There’s nothing like putting a record on the platter and letting it spin. (Playing music digitally is overrated!)

Believe it or not, Tamagotchi has been around for 25 years, which means that some of the original Tamagotchi users will be looking for updated versions for their own kids. The latest is the Uni which comes with Wi-Fi so that your creature can meet up with their pals online.
Tonieboxes are screen-free audio players that let even very young children pick their favorite tunes and stories. The basic collection can be augmented with content from Disney, Pixar, and Marvel, among others.

Not willing to put your iPhone in the hands of your baby? The Leapfrog Chat and Count Emoji Phone gives your little one plenty to occupy them, without running the risk that they’ll accidentally place an Amazon order, make a bank transfer, or message your boss.

If you’d prefer your child’s puppet to be more than just a plain old puppet, Disney has the Real FX alien (from Lilo & Stitch) which comes with animatronics.

This is not quite a tech toy, but Barbie – yes, the Barbie – now has a career set featuring two Barbies as chemists. Could there be a Barbie-pink Nobel Prize in their futures?

The Hajimari Ball is a high-tech boomerang spinner that’s billed as a way to get the kids off of their phones, tablets, and gaming devices. Plus it gets them outdoors. The Hajimari is a bit pricey, but there are plenty of flying-floating-droning-boomeranging orbs out there for the asking.

Another toy that touts itself as a means to get the kids up off the couch is the Steer Falcon Jet Fighter Stunt RC Airplane. It can fly! It can glide! It can hover! And it can perform a few stunts, too.

For a more down-to-earth remote-control vehicle, there’s the Turbo Drive RC which can be operated using hand gestures (or the usual remote-control console). There are plenty of gesture-driven remote-control cars, but this one seems to get uniformly high marks.

Most of the gifts on the Top Tech list are for older kids, but the Crabby Cruiser has infants in mind. When your baby’s on their tummy, propping themselves up, and getting ready to crawl, they’ll want to make a grab at this cruising crab, helping develop motor skills. Toddlers will have fun chasing after it, too.

Back to the older kids, the Spark Cubix is a Rubik’s Cube on steroids. While with the Rubik’s Cube, users “just” try to get the colors to align, the Spark offers multiple games and comes with the requisite flashing LED lights.

Nothing especially high-tech or science-y about the Bubblemi Blaster, but who wouldn’t want to be able to make thousands of bubbles, one shot at a time. The LED lights would be good for a nighttime event, but even without the lights, this one just looks like good old-fashioned fun.

The Black Falcon 4K is a small-ish drone that comes equipped with a 4K camera. One of the best things about it is affordability. (You may be able to find it on sale for a bit under $100.)

I go back and forth about tech toys for kids, sometimes confessing to a bit of nostalgia for the firetrucks and baby dolls of yore. But today’s world is tech-infused, and our kids and grandkids need to be fully equipped to make the most of that world. The beauty of most of the tech toys listed here is that they offer an option to kids who are glued to their screens pretty much all day.

Enjoy your holiday shopping!

 

What’s up with FPGAs?

As FPGAs have long been a technology used in a number of our system on modules, I read Steve Leibson’s recent post on EETimesThe State of the FPGA Union Is Uncertain – with great interest.

The uncertainty, as he sees it, is largely “thanks to tectonic upheavals with some key players.” But despite “tectonic upheavals,” he sees a positive outlook across the FPGA spectrum from the low end to the high end.

Where and why does Leibson see the uncertainty?

Well, last winter, Intel – which had acquired FPGA-maker Altera back in 2015 – decided to spin out its FPGA group into a new, somewhat independent, company. The new name has a back to the future feel. It’s doing business as Altera, An Intel Company. It may not be “An Intel Company” for long, as there’s an IPO on the horizon (2026).

Having Altera be Altera is a positive for the FPGA world, as it never really fit into the Intel processor world. Intel’s processors and tools, Steve writes, were built for making processors.

FPGAs are not processors. Instead, they are chock full of routing networks that connect sparse clumps of on-die logic and memory.

Construction-wise, FPGAs are nearly the complete opposite of processors, which rely on densely packed logic and minimally short routing networks to achieve their high performance. A semiconductor manufacturing process that is overly optimized for making processors—Intel’s bread-and-butter products—is necessarily suboptimal for making FPGAs.

Meanwhile, the other “top two” dominant player in the FPGA market, AMD, got into the FPGA world with its 2022 acquisition of Xilinx. But the biggest proponent of FPGAs at AMD, CEO Victor Peng has retired. Peng came along with the Xilinx acquisition, as he had been the CEO there for many years. Does this put FPGA-making at AMD “in a sort of limbo?” It remains to be seen whether AMD follows Intel’s lead and spins out of the FPGA business.

Altera is making some moves:

…modernizing its product line and is wisely migrating all three FPGA tiers—high-end, mid-range, and (ahem) cost-optimized—to a unified Agilex brand. That means the high-end Stratix, mid-range Arria, and low-end Cyclone and MAX FPGAs are at the end of their life cycles, from a development perspective. [Though these families will continue to be available for many years to come.]

Leibson doesn’t see all that much happening at the high end at either AMD or Altera, but does note that AMD/Xilinx tends to be ahead of Intel/Altera in coming out with new high-end products.

One likely explanation for this lag [between production announcements from AMD and Intel] is that processors get priority access to new process nodes at Intel, while AMD’s foundry of record for nearly the last 15 years, TSMC, provides customers with access to its latest process nodes as long as they will pay TSMC’s price.

There’s more happening at the lower end of the FPGA market, as both Altera and AMD have announced low-end FPGA production for 2025. But both AMD/Xilinx and Altera/Intel have a history of not prioritizing mid- and lower-end FPGA’s. This has enabled a couple of other FPGA vendors – Lattice Semiconductor and Microchip – to compete against the big guys.

Leibson’s bottom line is that the outlook for FPGAs is positive:

The entire FPGA industry benefits from the increased competition. There is a lot of life left in newer-but-not-leading-edge semiconductor processes that can be used for these more cost-sensitive FPGA market segments as the older manufacturing lines are depreciated. At the same time, the race to develop the next leading-edge process nodes provides ample opportunity for the high-end FPGA segment to grow, as well.

When it comes to FPGAs, what we’re seeing is very encouraging. We work in close partnership with Altera and have a few SOM families the use Xilinx as well. We are excited to see both manufacturers taking steps to grow board-level interest in their products, especially with Altera’s latest Agilex 5 FPGA, which we have developed a number of solutions around including a single board computer (MitySBC-A5E) and a family of system on modules (MitySOM-A5E). Customer adoption of SOMs continues to be on the rise, and the latest FPGAs deliver more performance with greater efficiency for customers developing advanced applications.

Gartner’s Tech Trends for 2025

One thing that I really enjoy about the end of the year is all the forecasts for the next year that start popping up.

Some headlines in the MIT Technology Review about a technology preview of the next century did catch my eye. But then I started thinking that this is so far out, it will impact the grandchildren of my great-grandchildren. So I thought I’d put that one aside for a later date – maybe one closer to the 22nd century, and focus on the almost here and almost now. Thus I landed on Gartner’s 2025 Top Strategic Technology Trends.

They grouped their trends into three categories – AI Imperatives and Risks, New Frontiers of Computing, and Human-machine Synergy – and outlined several sub-trends in each category.

AI Imperative and Risks
Under AI, Garner focused on agnetic AI, and on the challenges of governance, and figuring out how to combat disinformation. Agnetic AI pretty much means hands off. It’s a software approach that uses AI techniques and tools to make decisions and act on them in pursuit of goals – and to do so without much by way of human involvement. Agnetic AI has the potential to help organizations to achieve the promises of AI – greater productivity, freeing up workers to focus on higher-end tasks.

Empowering AI to act on its own leads, of course, to the need for governance platforms that can keep AI in check and “make sure AI is reliable, transparent, fair and accountable while also meeting safety and ethical standards.” This will be especially critical in industries that are highly regulated, but even non-regulated industries want to ensure that people’s privacy is maintained, and that there’s no unintended bias at play.

When it comes to AI, a problem we’ve increasingly the problem of disinformation, with more and more elaborate fraud schemes, crazier and crazier conspiracy theories that stoke fear and wreak havoc on society, and ever more elaborate (and convincing) attacks on businesses that undermine products and brands, wreck reputations, and manipulate stock prices. No wonder Gartner sees disinformation security coming to the fore.

New frontiers of computing
What does Gartner see for next year when it comes to computing? For one thing, cryptography in the post-quantum will become more and more critical, especially given that quantum computing will be deployed by criminals to decrypt stolen data, including data that was encrypted using traditional approaches. (Sophisticated cybercriminals have already planning for this future, with “harvest now, decrypt later” strategies, stealing even exceptionally well-encrypted data and hanging on to it with the assumption that, with quantum computing, they’ll be able to crack the code.

Gartner also sees ambient invisible intelligence in the future. This is “the widespread use of small, low-cost tags and sensors to track the location and status of various objects and environments.” The amount of data available for analysis will continue to grow exponentially, and the “real-time visibility” it provides will have a major impact on a range of industries: supply chains, healthcare, retail.

Demands for sustainability will grow, especially as AI is such a tremendous driver of energy consumption.

While conventional processing improvements are reaching their limits, new computing technologies, such as graphics processing units (GPUs), neuromorphic computing and quantum computing, are expected to deliver the substantial energy-efficiency gains needed in the next five to 10 years.

Also up will be hybrid computing, where all sorts of technologies “— such as CPUs, GPUs, edge devices, ASICs, and neuromorphic, quantum and photonic systems — “ are combined to solve especially thorny problems.

Human-machine synergy
It’s good to keep in mind that, behind all the AI, beyond all the quantum computing, there are human beings, and Gartner foresees “enhanced interactions” between the physical and virtual worlds; robots being an increasing part of our every day lives; and “technologies that directly influence cognition and performance.”

Spatial computing will be getting more attention as augmented and mixed reality, along with AI technologies in general, will make it possible for “users to interact with [the physical world] in an immersive, realistic and intuitive experience.”

Robots will become more skilled, with polyfunctional robots able to take care of multiple tasks. E.g., warehouse robots will be able to pick, pack, and transport goods. And, in the final trend Gartner identified, neurological enhancement, “using technologies that read and decode brain activity and optionally write to the brain,” will improve human cognition, which will hopefully allow us to keep up with those polyfunctional robots.

 


Source of image: Vistage

Trick or Treat Tech

It’s that time of year, when the neighborhood starts getting all spookied up for the little ghosts and princesses, the little Blueys, the Barbies, the Spidermen, who’ll be ringing our doorbells next week.

Each year, Halloween yard decorations seem to keep getting more elaborate and more technology based – not to mention more expensive. And if you’re a Halloween yard-decorator, you’ve probably had your decorations up for a few weeks already. But if you haven’t, it’s not too late! So we took a look at what’s out there.

Ever since they introduced us to Skelly, the 12 foot skeleton, in 2020, Home Depot has been coming up with enhancements and variations on a theme. This year’s version of Skelly now has glowing LED eyes that can be customized. Theoretically, this is so that Skelly can be used for other holidays, although I don’t know what other holidays a 12-foot skeleton applies to, but that may just because I’m lacking in imagination here. If you’re the type who doesn’t mind having a 12-foot skeleton outside your front door year-round, Home Depot offers a Santa costume for Skelly, and the Skelly LCD eye kit provides a range of remote-controlled eyes , including Valentine’s Day hearts and fireworks for the Fourth of July You If you’re looking for a more technically advanced skelly, there’s an animatronic version available.

If you’ve been there/done that with Skelly, or just want a companion piece for it, Home Depot has a 12-foot Animated LED Levitating Reaper that appears to float and utters a number of creepy phrases. And if 12-foot is a bit too tall for a yard decoration, Costco has a 7-foot flannel-shirt wearing werewolf. It comes with LCD eyes and sound effects, and all of it’s activated by a motion sensor. (Let’s hear it for embedded systems!) You can go shorter still at Lowe’s, with a 6-foot tall animatronic deep-sea diver from Lowe’s, that like its larger kinfolk, has glowing eyes, makes noise, and moves around.

By the way, all of these items may cut into your candy budget, as they’re all in the $300+ range. Boo!

Of course, $300+ is nothing compared to the Distortions Unlimited Scare Wolf. “The frightening werewolf animatronic lunges at unsuspecting bystanders. A hose in its mouth blasts air for a viscerally interactive experience.” All this and more for $5,250, which makes the $340 Costco Werewolf seem like a bargain.

You can take your scary creatures indoors for a lot less money, with digital downloads of all kinds, including AtmosFX’s talking scarecrows.

There are all sorts of other techie Halloween decorations out there. LED! Animatronics! Holographic! Witches! Bats! Eyeballs!

If you’re into multiple holidays, however, a good way to simplify your decorating life is to install permanent outdoor LED lights and change the color scheme. The eufy E22 lighting system is AI-powered. You can choose the “occasion and the mood,” and the algo will do its thing and create a new look and feel for whatever holiday and season you want to celebrate.

Halloween decorations have certainly come a long way from the days of the pumpkin carved into a Jack O’ Lantern and set out on the front stoop. (Quality varied with the skill level and age of the carver!) Or when decorating meant Kleenex ghosts dangling in the windows.

Anyway, it’s time to make sure we have enough candy for the kids that’ll come by. (With a bit left over for ourselves, of course.)

Happy (Technical and Non-technical) Halloween.

 

Sources: CNET and The Gadget Flow

Words to the wise

Those of us who’ve been around the product development world for a while know the stages of product development pretty much by heart. While some aspects of the process have changed over the years –  think rapid prototyping – the core elements have stayed the same. From identifying a need to coming up with ideas, from specification through prototyping, from testing through refinement and onto release, there are certain steps that product development teams follow.

But as engineers, how much thought do we give to what comes afterward, once the products we’ve worked on are out in the market?

Sure, we get to respond to requests for tweaks and new features, and have to make fixes when we’re the recipient of unwelcome (but never all that surprising) bug reports. Or, in this day and age, a very public complaint about some failure or another that’s appeared on social media.

But how much thought does a typical engineer give to how they’re working with customer support – the crew that’s the first line of defense when issues arise?

The answer, if we’re being honest, is probably not all that much.

So it was good to see a reminder from Chris Coleman, writing recently on embedded.com, that addresses “how customer support and engineering teams can work better together.”

Coleman writes from the perspective of someone with considerable engineering experience in the embedded world, largely in the wearable space. He went on to found a company that has a platform focused on the “observability” of embedded IoT and edge devices, monitoring, analyzing, and responding to crash data. So he has a foot in both the development and ongoing support worlds. And, naturally, since it’s what his company does, it’s no surprise that he argues for engineering to be proactive by monitoring devices, rather than waiting for often isolated, one-off bug reports/support tickets from customer support coming their way. Still, the argument he makes is pretty compelling.

He begins his piece by noting that, despite all the testing in the world, it’s just not possible to anticipate every last way that your product will be used, or all the “environmental conditions” it will meet up with. We see this with our SOMs all the time, as more and more customers from diverse industries, with new applications adopt them. Yes, you can set out the use cases and conditions that your product will support, but even within those boundaries, the unanticipated can happen. And it can be almost impossible to replicate exactly what happened leading up to a problem occurring.

Next, he outlines what customer support does to respond to problems. We all know the continuum here, from restart the device to send it back to have us take a look at it. And we all know that customer support works with tiers with respect to the criticality of an issue, and that they group problems in buckets.

These reports will then make their way to engineering teams. Being on the other side of receiving this information, it’s not super actionable. Often very different underlying problems manifest in the same way. It’s hard to know what exactly to fix and what changes will actually drive down the bulk of reports.

Over the years I’ve talked to so many support teams that see engineering as where bug reports go to die – very rarely do any resolutions come back. I’ve also seen many engineering leaders find the support requests inactionable – it’s very hard to go from a summary like “80% of issues are around connectivity” to actually fixing the problem.

Not much to say here, other than “ouch.” Or maybe been there, done that.

As I said earlier, Coleman is putting forward the case for adopting his company’s product. Nothing wrong with that. And nothing wrong with suggesting that engineering should work more closely with customer support. As an ISO 9001:2015 certified company, Critical Link has processes in place that ensure we close the feedback loop from customer support back to engineering. This helps our team to be vested in the entire product lifecycle, and ensures continuous improvement in our designs. From Coleman’s perspective, he advocates having embedded engineers begin to utilize practices that focus on tracking core vital signs with respect to device reliability. This will give engineering a better handle on problems, and provide support with a better, more informed means of communicating with customers. And no more tickets that never get closed out.

My main takeaway from reading this piece is that engineering and support can never operate independently of each other, and if there are approaches (and tools) that make products more supportable and customer support more capable of truly supporting their customers, I’m all for it. Consider these words to the wise.


Source of image: Buyer Foresight

Smartphone app processors

First, a confession: I didn’t read the full report. It would have cost me nearly €7,000 (or nearly $8,000 US). But fortunately, Don Scansen apparently did, and wrote a good but too brief summary of APU – Smartphone SoC Floorplan Comparison 2024, the Yole Group’s study of smartphone application processors that was published in the EETimes last month.

The study focused on the leading device manufacturers, and on the designs that were the most current as of the study date. (So, there may already be new products out there, as this is a dynamic and very competitive market.)

The products included in the study were:

  • Apple A17 SoC in iPhone 15 Pro
  • Google Tensor G3 SoC in Pixel 8 Pro
  • HiSilicon Kirin 000s SoC in Huawei Mate 60 Pro
  • MediaTek Dimensity 9300 SoC
  • Qualcomm Snapdragon 8 Gen 3 SoC
  • Samsung Exynos 2400 in Galaxy S24 SM-S921U1

Scansen noted that there is one key feature that all six design groups have in common: the processors all rely on ARM designs. After that, things start to diverge. Apple uses ARM IP, “but assembles the logical functions into their own physical designs,” an approach that makes sense since Apple is the only company on the list that has responsibility for both hardware and software in their products. The other design teams deployed hard macros from ARM.

The A17 design stands apart in other details, as well. The standard approach to CPU architecture is to have performance cores for the most demanding tasks and smaller efficiency cores to operate with lower power consumption whenever appropriate. The Apple A17 performs the most intense software operations with the lowest core count—just a dual-core design.

Apple holds the position at one end of the design spectrum. MediaTek is at the other. Their Dimensity 9300 uses “the most powerful CPU.”

This APU contains four of the supersize cores that other designs tend to use only as a single main core supported by smaller high-performance cores. The Dimensity 9300 has a complete high-performance quad-core section of Arm Cortex-X4 cores. For power-saving functions, MediaTek prefers the Cortex-A720 core (typically used for high-performance operations) rather than one of the efficient Arm variants.

Scansen provides considerable discussion on the impact on HiSilicon Kirin, a Chinese design company, and SMIC a Chinese domestic fab (which works with HiSilicon) of sanctions that have been imposed on the use of Chinese foundry tools and advanced lithography.

SMIC built the Kirin 9000 on a lagging node (7 nm) compared to 3 nm (TSMC for Apple) and 4 nm (Samsung for its own and Google processors, and TSMC for Qualcomm and MediaTek). That might suggest the HiSilicon design would be among the largest chips. However, MediaTek’s processor has the biggest die despite being built on the much more advanced TSMC 4-nm platform. Although the TSMC 4-nm process increased transistor density, MediaTek’s design with four of the largest Arm X4 cores necessitates expansive use of silicon.

He then explores die size among the different products. (Most are roughly comparable.) Scansen’s focus here is largely to contrast Apple (which builds using ultra-advanced foundry technology), with HiSilicon (which sits on “the most trailing edge of the group”). He continues the contrast between these two by pointing out that the Apple A17 “includes about twice as many transistors as the Kirin chip.”

Scansen devotes a lot of space to the HiSilicon Kirin 9000 (more on the issues they had when they had to retreat from use of made-in-China technology), and I wish that he’d provided more information on the other products on the list. But, as he says:

There are many twists and turns, as well as chip partitioning and placement tradeoffs when comparing these top-tier smartphone processor designs. A complete understanding requires more time and space.

I hope that Don Scansen finds that time and space. Because, although nearly €7,000 (or nearly $8,000 US) for the Yole Group report is too rich for my blood, I’d be very interested in hearing more about what’s in it.


Image source: Hatch.

Supershoes!

Although I’ve done some running over the years – some Spartan races, a 5K or two – I am by no means a serious runner. Still, I do have a lot of admiration for runners – especially distance runners – and read with great interest a fascinating article by Jonathan Rosen on the latest high tech running shoes, an article that appeared in the MIT Technology Review this past June.

The article centered on runners from Kenya, and that’s no surprise. As Rosen notes, “Kenya-born athletes accounted for 28 of the event’s all-time 50 fastest men and 17 of its 50 fastest women.” The official marathon record (2:00:35) was set by Kelvin Kiptum last October, and the unofficial record – and only sub-two-hour marathon – was set by Kiptum’s fellow countryman Eliud Kipchoge in 2019. (Time 1:59:40.2.) Kiptum was hoping to run a sub-two-hour race when, sadly, he was killed in an accident last winter.

Kenyans have achieved their stunning success for a number of reasons. The tribes that make up the majority of the country’s top runners share physical characteristics that make for an “energy-efficient gate.” These characteristics include “thin lower legs, long Achilles tendons, and a high ratio of leg length to torso.” Kenyan children are very active – they run places –   and they’re running at high altitudes that “help boost aerobic capacity.” Then there are the financial drivers. While Kenya is more industrialized and prosperous than many other African nations, many people live in poverty, and running is viewed as a way out.

But one key to the record shattering is the “supershoes” that competitive runners – and not just those from Kenya, of course – are wearing.

It was interesting to read that running shoe “technology” changed little between 1908 (when rubber heels made their Olympic debut) and 1975 when ethylene vinyl acetate (EVA), and “air-infused foam” made running shoes that were lighter and bouncier. And for decades after, most running shoe improvements were for the most part EVA tweaks. And then, in 2017, Nike introduced the Vaporfly, which had a curved carbon-fiber plate and “a thick midsole material made from a foamed polymer known as polyether block amide, or PEBA.” That’s what puts the pep in a runner’s step (and helps runners suffer fewer injuries and recover more quickly from a grueling race).

And since 2017, this has been the name of the supershoe game.

The most crucial feature is the (often proprietary) foams that are used to construct parts of the sole. These absorb the impact of the foot and return energy from each foot strike back to the runner. Some use other features, like the orange “air pod” in the Nike Alphafly 3, for an added bounce.

Bounciness alone would not provide much advantage—today’s foams are so soft and thick (World Athletics allows up to 40 millimeters in competitions) that without additional support they would make the feet highly unstable. To give the shoes structure, manufacturers add rigid components like carbon-fiber plates or rods, typically sandwiched between layers of foam.

These rigid parts and foams are combined with wafer-thin mesh uppers to create shoes that are increasingly ultralight: the Adidas Adizero Adios Pro Evo 1, released in 2023, weighs just 4.9 ounces (measured in the men’s size 9). Lighter shoes also reduce the energy expended with each stride—enabling runners to move at a given pace with less effort. The Adidas Adizero Adios Pro Evo 1 was designed to be worn just once.

Interestingly, there’s research “that some runners get a greater boost from the technology than others, depending on their biomechanics.” And even those who are most in favor of supershoes (not everyone is) don’t want to take away the from the “talent, training, drive, and mental toughness” of the world’s most successful runners. Training has also improved over the years, and “new hydrogel-based sports drinks…make it possible to digest more calories during races.” (Kiptum’s coach had “estimated that the shoes improved Kiptum’s marathon time by a minute, or perhaps a little more.” Which is, of course, pretty significant, especially in the major races with large purses where split-second timing can translate into tens of thousands of dollars.)

As for Kenyan runners, the supershoes are to some degree a mixed blessing. For the top runners, it makes a performance difference. For those trying to make their mark, it works against those who can’t afford them.

Anyway, even though I’m not that much of a runner, and running shoe technology isn’t the sort of technology I’m involved with, I found the story about supershoes a fascinating one.

Taking industrial control systems offline

Much has been written about the Industrial Internet of Things (IIOT). And we’ve even contributed a bit of what’s been written ourselves, as we did with a June post on the IIOT. As has been widely noted, one of the main challenges facing the IIOT is its potential vulnerability to cyberattacks. As it now appears, it’s also plenty susceptible to problems that have nothing to do with security incursions by bad actors. We saw this July when a software update to Microsoft-based system – an update being made by CrowdStrike, a cybersecurity company, of all things – went awry. The resulting outages were global, and impacted a number of critical industries, including banking, healthcare, and the airlines.

The CrowdStrike outage will result in increased scrutiny of vulnerable systems, and maybe even some rethinking about the wisdom of uber-connectivity.

A piece on embedded.com in late June anticipated just such rethinking when it comes to the industrial control systems (ICS) that govern so many of our most important industries. In his post, Richard Kanadjian makes an argument for taking ICSs – a prime attack target – offline, and makes some suggestions for going about it.

Kanadjian details the “slew of warnings and security alerts” that have been emanating from various organizations of late. In May, Rockwell Automation – a leading ICS provider issued an alert suggesting that its customers remove ICSs that were not designed for “online exposure” from the Internet. Earlier on, beginning in 2020, both the National Security Agency (NSA) and CISA (Cybersecurity and Infrastructure Security Agency) have put forth regular advisories pointing out just how vulnerable internet-connected ICs are and, along with NIST (the National Institute of Standards and Technology) recommend “air-gapping,” which:

… involves physically isolating ICS from any internet connections, effectively eliminating the risk of remote cyberattacks. When a system or server is completely air gapped, it has no communication at all with the outside world, making it invisible and inaccessible to hackers. (Source: embedded.com)

This raises the question about how to manage software updates, patches, and data transfers, which have increasingly been automated and/or cloud-based.

As the recent CrowdStrike debacle illustrated, there’s much that can go wrong with updates that have nothing to do with security, but security remains a concern. Here the approach Kanadjian recommends is:

…hardware-encrypted mobile storage solutions, such as OS-independent, hardware-encrypted USB or SSD drives, [to] securely transfer software updates and data. These drives require a PIN for access, ensuring that only authorized personnel can transfer data. Typical encryption software is not enough to maintain security, as many ICS machines may not be able to run software encryption like AES-256, making a self-contained solution necessary. Storage drives, designed from the ground up to be data protection drives with a dedicated secure microprocessor, are compatible with various ICS operating systems, providing a versatile and secure method for updates, and meeting all criteria for the CIA Triad cybersecurity requirements.

Note that Kanadjian works for Kingston, which is not a neutral party here. This is Kingston’s business. The suggested cannot, however, be discounted.

The systems we rely on for national security, for our economic and physical well-being, are vulnerable to attack. If something as seemingly trivial as a failed software update can result in widespread flight cancelations, delayed medical procedures, and lack of access to banks, imagine the havoc that could be wreaked if a deliberate attack by a bad actor was made on our most critical systems.

We take so much that happens behind the scenes for granted, but when it comes to our industrial control systems, we obviously need to be hyper-vigilant.

 


Image Source: Driver Easy

AI-driven robots that can help folks live more independently? Bring it!

A couple of weeks ago, I posted about attempts to help alleviate one of the downsides of AI: its power consumption, and the environmental impacts that comes with.

What I didn’t mention in that post is that there are small ways in which we can help reduce AI-related power consumption, and that’s by not buying into any “AI-for everything” madness, in which AI is used even when it’s not needed and/or adds limited value.

As consumers, do our home appliances have to be all that smart? Do we really need AI to tell us what’s in the fridge when we can just open the door and look? Do we need to deploy AI for every Google search we do? Shouldn’t we be willing to “wait” an extra second for a non-AI search that may well yield superior results, by the way. (Public Service announcement: if you don’t want Google to produce an AI Overview for every search you do, use the Web option when you search.)

But there are plenty of applications where AI can and should be used in the home. And one was chronicled last spring in MIT’s Technology Review.

More than twenty years ago, Henry Evans – then only 40 years old – suffered a major stroke and ended up a quadriplegic who was unable to speak. Over the years, he was able to use his eyes and a letter board to communicate, but in most day-to-day situations, Henry has to rely on caregivers.

Then robotics came on to the scene. Sort of.

In 2010, Henry saw a demo of a primitive “metal butler,” and asked himself why something like that wouldn’t work for him.

There was a solid reason why not. While engineers have made great progress in getting robots to work in tightly controlled environments like labs and factories, the home has proved difficult to design for. Out in the real, messy world, furniture and floor plans differ wildly; children and pets can jump in a robot’s way; and clothes that need folding come in different shapes, colors, and sizes. Managing such unpredictable settings and varied conditions has been beyond the capabilities of even the most advanced robot prototypes.

But thanks to AI, that may be about to change, giving robots the opportunity to advance beyond the skills that are driven by purpose-built software and to acquire new skills and figure out new environments faster than they ever could before.

Henry Evans has already been working with experimental robots that are letting him take care of tasks like brushing his hair. “Stretch,” the robot Henry is currently working with – the brainchild of Georgia Tech professor Charlie Kemp – goes beyond specific-purpose tasks, like hair-brushing, and lets users “plug in their own AI models and use them to do experiments.” As Stretch learns more, it can do more.

With AI-software-powered robots, robots will be able to acquire new skills automatically rather than have to solve each problem independently and by having to plot each element in “excruciating detail.” Not that these painstakingly acquired skills aren’t impressive. Who hasn’t marveled at a video of humanoid (or dog-oid) robots – climbing stairs, boogeying, opening doors? Now research is moving beyond pure physical dexterity and are now experimenting with “building ‘general purpose robot brains’ in the form of neural networks, and tapping generative AI in ways that go beyond “the realm of text, images, and videos and into the domain or robot movements.” This will allow robots to quickly adapt to new environments and quickly learn new tasks.

The cost (and size) of robots will come down, their utility will go up, and for folks like Henry Evans, the world will open up. It already is. While Stretch is imperfect – it’s buggy and bulky – it’s a declaration of independence. “All I do is lay in bed, and now I can do things for myself that involve manipulating my physical environment.” These include playing with his granddaughter, holding his own hand of cards, and eating a fruit kabob.

I’ve always maintained that one of the very best things about being an engineer is doing work that really does improve people’s lives.

AI-driven robots that can help folks live more independently? Bring it!

 

_________________________________________________________

Source of Image (Henry Evans giving his wife Jane a rose): IEEE Spectrum