When it comes to repairing human bodies, there’s one major difficulty: spare parts are hard to come by. It’s simply not possible to buy a knee joint or a new lung off the shelf.
At best, doctors and surgeons have made do with transplants from donors where possible. However, these are always in short supply, and come with a risk of rejection by the patient’s body.
Needless to say, the views we get through modern lenses are a lot more realistic. So how did we get from simple magnifying systems to the complex lens systems we see today? We start with a quick journey through the history of the camera and the lens, and we’ll end up with the cutting edge in lens design for smartphone cameras and VR headsets.
At the risk of dating myself, I will tell you that grew up in the 80s — that decade of excess that was half drab and half brightly colored, depending on where you looked, and how much money you had for stuff like Memphis design. Technology seemed to move quickly in almost every aspect of life as the people of the Me decade demanded convenience, variety, and style in everything from their toilet paper (remember the colors?) to their telephones. Even though long distance cost a fortune back then, we were encouraged to ‘reach out and touch someone’.
A Healthy Fear of Bears
Looking back, it’s easy to see how all that advanced technology and excess filtered down to children. I may be biased, but the 80s were a pretty awesome time for toys, and for children’s entertainment in general. Not only were the toys mostly still well-made, even those that came in quarter machines — many of them were technologically amazing.
Take Teddy Ruxpin, which debuted in 1985. Teddy was the world’s first animatronic children’s toy, a bear that would read stories aloud from special cassette tapes, which moved his eyes and mouth along with the words. One track contained the audio, and the other controlled three servos in his face.
I remember watching the commercials and imagining Teddy suddenly switching from some boring bedtime story over to a rockin’ musical number a là the animatronic Rock-afire Explosion band at ShowBiz Pizza (a Chuck E. Cheese competitor). That’s the kind of night I wanted to be having.
The current lineup of the Rock-afire Explosion. Image via Servo Magazine
Which brings us to KC Bearifone, an animatronic teddy bear telephone. Honestly, part of the reason I bought the Bearifone was some sort of false nostalgia for Teddy. The main reason is that I wanted to own a Teleconcepts unit of some kind, and this one seemed like the most fun to mess around with. A robot teddy bear that only does speakerphone? Yes, please.
It’s fair to say that there’s really no phase of spaceflight that could be considered easy. But the case could be made that the most difficult part of a spacecraft’s journey is right at the very beginning, within the first few minutes of flight. At this point the vehicle’s booster rocket will be fighting with all its might against its own immense propellant-laden mass, a battle that it’s been engineered to win by the smallest of margins. Assuming the balance was struck properly and the vehicle makes its way off of the launch pad, it will still need to contend with the thick sea-level atmosphere as it accelerates, a building dynamic pressure that culminates with a point known as “Max q” — the moment where the air density imposes the maximum structural load on the rocket before quickly dropping off as the vehicle continues to ascend and the atmosphere thins.
Air-launched rockets avoid flying through dense sea level air.
While the vast majority of rocket launches have to contend with the realities of flying through the lower atmosphere, there are some exceptions. By launching a rocket from an aircraft, it can avoid having to power itself up from sea level. This allows the rocket to be smaller and lighter, as it doesn’t require as much propellant nor do its engines need to be as powerful.
The downside of this approach however is that even a relatively small rocket needs a very large aircraft to carry it. For example, Virgin Orbit’s LauncherOne rocket must be carried to launch altitude by a Boeing 747-400 airliner in order to place a 500 kg (1,100 lb) payload into orbit.
But what if there was another way? What if you could get all the benefits of starting your rocket from a higher altitude, without the cost and logistical issues involved in carrying it with a massive airplane? It might sound impossible, but the answer is actually quite simple…all you have to do it throw it hard enough.
The war in Ukraine has upset the global food market, and the surprising reason is not that Ukrainian wheat isn’t being harvested, but rather that it can’t leave the country. With Russia blockading sea ports, the only way out for Ukrainian grain is by train. And this exposes the long-hidden patchwork of railway tracks and train standards: trains can’t simply cross the border from Ukraine to Poland on their way to a sea port because the tracks don’t match.
Even beyond the obvious issues of connecting differently sized physical railway tracks — the track gauge — there are different signaling systems, different voltages for electrical trains, different loading and structural gauges, and so on. In Europe today, the political history of the past few hundred years can still be traced back using its railroads, with some parts of the European Union still on 1,520 mm Soviet-standard gauge, rather than the 1,435 mm Standard Gauge, which is also known as Stephenson Gauge, European Gauge, etc.
These complications explain why for example with the current war in Ukraine its railways into the rest of Europe aren’t used more for transporting grain and other cargo: with Ukraine using 1,520 mm gauge, all cargo has to be transferred to different trains at the Ukraine-EU border or have bogies swapped. Although some variable gauge systems exist, these come with their own set of limitations.
In light of this it’s not hard to see why standardizing on a single international or even European track gauge is complicated due to having to replace or adapt all tracks and rolling stock, even before considering the aforementioned voltage and signaling differences. All which may lead one to wonder whether we’ll ever see a solution to this historically grown problem.
Four years ago when the idea of a pandemic was something which only worried a few epidemiologists, a group of British hardware hackers and robotic combat enthusiasts came up with an idea. They would take inspiration from the American Power Racing Series to create their own small electric racing formula. Hacky Racers became a rougher version of its transatlantic cousin racing on mixed surfaces rather than tarmac, and as an inaugural meeting that first group of racers convened on a cider farm in Somerset to give it a try. Last weekend they were back at the same farm after four years of Hacky Racer development with racing having been interrupted by the pandemic, and Hackaday came along once more to see how the cars had evolved. Continue reading “How Far Can You Push A £500 Small Electric Car; Four Years Of The Hacky Racer”→
“With the power of edge AI in the palm of your hand, your business will be unstoppable.”
That’s what the marketing seems to read like for artificial intelligence companies. Everyone seems to have cloud-scale AI-powered business intelligence analytics at the edge. While sounding impressive, we’re not convinced that marketing mumbo jumbo means anything. But what does AI on edge devices look like these days?
Being on the edge just means that the actual AI evaluation and maybe even fine-tuning runs locally on a user’s device rather than in some cloud environment. This is a double win, both for the business and for the user. Privacy can more easily be preserved as less information is transmitted back to a central location. Additionally, the AI can work in scenarios where a server somewhere might not be accessible or provide a response quickly enough.
Google and Apple have their own AI libraries, ML Kit and Core ML, respectively. There are tools to convert Tensorflow, PyTorch, XGBoost, and LibSVM models into formats that CoreML and ML Kit understand. But other solutions try to provide a platform-agnostic layer for training and evaluation. We’ve also previously covered Tensorflow Lite (TFL), a trimmed-down version of Tensorflow, which has matured considerably since 2017.
For this article, we’ll be looking at PyTorch Live (PTL), a slimmed-down framework for adding PyTorch models to smartphones. Unlike TFL (which can run on RPi and in a browser), PTL is focused entirely on Android and iOS and offers tight integration. It uses a react-native backed environment which means that it is heavily geared towards the node.js world.
