In Australia, you currently need a recreational helicopter pilot’s license to fly this machine – but US owners might not. “It is possible to fly the Pegasus without a pilot’s license,” reads the company website, “if you have successfully completed the FAA private Pilot written examination and have also completed the company-mandated vehicle familiarisation and operator training programs.”

Although somewhat tangential to this community’s intent, I decided to verify this claim, since licensing and registration of an air-and-land machine is likely to be a top question.

Regarding United States aviation, everything revolves around certificates. For individuals, a certificate is a license to pilot or operate particular types of aircraft. For machines, a certificate is the authorization of airworthiness for a particular type of aircraft, allowing it to be used in US airspace. Setting aside the road registration quandary – which is regulated on a state-by-state level – I will focus only on the aviation aspects, since those are controlled mostly by federal law.

For pilot licensing, there are a fair number of those, ranging from the Student Certificate to get into the air under instruction, to the traditional Private Pilot Certificate/License (PPL) for flying solo or cross country, to the Air Transport Certificate to fly paying customers commercially.

Given the description from the company, I would guess that they meant the Light Sport Certificate, which is an abbreviated pilot license to fly smaller aircraft with up to one passenger and weighing less than 1320 pounds (600 kg). This certificate does not require the thorough medical exam of a PPL, if the holder also has a US Driving License. This might sound a bit weird, since why would driving an automobile be indicative of sufficient health to not crash an airplane, but it’s a balancing act given the restrictive set of aircraft types that can be flown with a Light Sport certificate.

So the company’s statement is vaguely, mostly correct, if they meant that the strict requirements of a PPL can be avoided, by instead applying for a Light Sport certificate. Although this still requires 20 hours of flight training beforehand.

As for certificating the aircraft itself, this is fairly straightforward, since the manufacturer just needs to declare that they meet all the requirements in 14 CFR § 1.1. The FAA would then grant the type certificate, allowing this aircraft into the national airspace. The owner would then need to register the aircraft by presenting the type certificate, and then receive a tail number (aka N registration) to attach to the aircraft.

So if a person is in possession of a valid Sport Light certificate, an N registration with the FAA, and this helicopter/go-kart, they should be good to take off, right? Well, mostly, but with a substantial number of caveats.

Firstly, certificated aircraft – even for the Light Sport or Experimental categories – is still strict, and any modifications to the airplane that deviates from the regulations can invalidate the type certificate. Even just basic maintenance must be performed by a certificated mechanic, as well as the mandatory annual inspection. While one could obtain that additional certificate to maintain one’s own Light Sport aircraft, the inspection must take place at a certificated inspection location, which is probably somewhere else.

And then there are the operational limitations. This is less of a problem with the machine itself, since the limits in 14 CFR § 1.1 simply won’t impose a restriction. Rotorcraft don’t usually fly high enough that they need pressurization, and this machine only comes in a one- and two-seat variant. The real issue is the pilot’s license limitations.

With only the Light Sport pilot license, it is Visual Flight Rules (VFR) only. So no flying at night, through or over the clouds, nor into inclement weather where visibility is below 3 miles (5 km). And only Class E and G airspace are permitted, unless having obtained additional endorsements to operate near airports and to communicate using the radio.

This page describes US airspace circa 1997 for ultralights (a category of very light flying machines that the FAA doesn’t even require type certification for), but it does underscore the complexity in determining what airspace is of which class. Ultralights and LSA do tend to use the same airspace, though.

Perhaps you could argue this go-kart/helicopter is only meant for flying close to the surface, so the roughly 700 ft upper bound for Class G airspace would be perfectly sufficient. Maybe. But most urban areas have an airport within 10-15 miles (16-25 km), which may displace the Class G airspace to a 0 ft upper bound, meaning the Class G space is non-existent. That really cuts against using this as an urban commuter.

But maybe it’s meant for rural/suburban commuters who don’t necessarily go into the urban core but travel around it. But now we’ve reduced this machine to a one-trick pony.

Do I think this is an intriguing machine? Yes, absolutely; I wouldn’t have hashed out this long comment if it weren’t at least some food for thought. But do I think this is the future for the masses? Definitely not.

The process to get this legally flying is sufficiently involved that it will never see mass appeal. And I will not even entertain the notion that flying regulations should be relaxed to accommodate this novelty, since we already have an example where lax licensing of heavy machines and grossly insufficient operator training causes thousands of deaths per year needlessly: automobiles.

Most people riding bicycles already deploy their retractable landing gear when preparing to touch down: they’re called legs.

And while maybe there’s an argument to be made that a properly-adjusted bicycle might not allow both feet to touch the ground while still seated, this contraption has some serious limitations. The most egregious is that landing gear – like on aircraft – are only effective when fully extended. But whereas all runways are laterally level, bicycling surfaces offer no such guarantee.

Imagine riding this cargo bicycle on a curve which has a cant (aka superelevation, or banking, or cross slope), then slowing down for a pedestrian or even stopping. A two-wheel bicycle naturally finds the lean angle to balance the centripetal force against the gravitational force, irrespective of the relative angle to the ground surface. But fully-extending landing gear forces the bicycle to be perpendicular to the ground surface, against the mandatory lean angle to balance the forces. And there doesn’t seem to be a provision for partially extending one side of the landing gears to account for the cross slope. If the perpendicular arrangement puts the center of gravity beyond the stubby extended wheels, the bike will fall over.

Tricycles will also force the same perpendicular alignment, but: 1) does so at all times, on straightaways and curves, so this isn’t a surprise to the rider, and 2) tricycle wheels are set wider than these stubby “landing gear” wheels, which is important if the cargo load is substantial (or heavy or too high) and could topple the bicycle when stopping on a curve.

This contraption is akin to bicycle training wheels: too narrow to provide actual safety benefits, while also being outright detrimental towards developing the motor skill necessary to pilot a cargo two-wheeler with its unique qualities.

Absolutely, it is essential to always run the numbers. I was once offered a sizable rebate if I accepted a non-0% car loan, but no rebate if I paid cash or had my own financing. Since their loan had no early-repayment penalty – and I demanded this in writing – I accepted their loan and paid it off upon the first statement.

My suspicion is that that sort of offer was to boost the commissions earned by the loan brokers, rather than to move cars. Or maybe both. Who knows.

0% interest offers show up fairly frequently in the USA, often as general-purpose credit cards, or for car or furniture payments, in addition to the many buy-now-pay-later services that allow financing almost anything. However, the motives for offering 0% are slightly different for each of these products.

But answering the question directly, a 0% offer is beneficial if you were already going to make the purchase and would finance it. Cheap credit makes it easy to overspend, since the payments will be “tomorrow’s problem”. For people who can afford to pay for something in full, it might still be beneficial to finance with 0% just to conserve cash on hand. But the tradeoff is having to service the debt with regular payments; missing one payment can cause the debt to resume at an exorbitant rate. It takes a decent amount of financial discipline to make a 0% offer work in your favor.

Going back to why 0% offers even exist, I’ll use furniture and cars as they’re the historic examples. Furniture is expensive, whether it’s a sectional sofa or a queen-size bed with frame and storage. There’s also a sizable markup for furniture, and competition between furniture stores is strong. Thus, to help entice people to buy furniture, sellers will offer 0%, outsourced to a loan company, with the loan subsidized by some of the profit margins.

For cars, the equation is slightly different. Sure, cars are an order of magnitude more expensive, but that also means the opportunity cost for dealers to offer 0% is correspondingly larger. Instead, 0% financing for cars is almost always subsidized by the manufacturer, not the dealers. This is a financial and business strategy that allows a car company to create more sales in a given quarter, if perhaps they need to meet certain year-end targets but are reluctant to reduce their list prices.

0% car loans induce more sales fairly quickly, but will draw on the company coffers in the years to come, because the loan company still wants their cut to be paid by someone. Consumers will usually benefit from these offers, as it’s rare for people to buy a new car outright.

It’s my opinion that if a car company has to subsidize loans to move their product, that’s a tacit admission that their product is wrongly priced or the competition is better. I would take this into consideration, although it wouldn’t necessarily carry the day when considering a purchase. After all, car payment interest is not insignificant.

Agreed. When I was fresh out of university, my first job had me debugging embedded firmware for a device which had both a PowerPC processor as well as an ARM coprocessor. I remember many evenings staring at disassembled instructions in objdump, as well as getting good at endian conversions. This PPC processor was in big-endian and the ARM was little-endian, which is standard for those processor families. We did briefly consider synthesizing one of them to match the other’s endianness, but this was deemed to be even more confusing haha

It never ceases to amaze me how prolific PowerPC/PowerISA was (still is?) in the embedded space

A quick “rule” is to see how old the word/concept is. “plaintiff” would have existed almost as long as the English legal system came into being, or probably even older to the court of Assizes pre-12th century.

Whereas firefighter as a profession might have only become a word after the establishment of fire departments by insurance companies, which I think might have been a 19th century development.

This entire series by Cathode Ray Dude is a wonderful dive into the world of PC boot sequence, for the folks interested in a touch of embedded architecture. His delivery is also on-point, given the complexity and obscurity of the topics.

From this video alone (41:15):

The way this worked was: they installed Xen hypervisor on your PC, put Hyperspace in a VM and Windows in another. Now, you either know what a VM is – and I don’t need to explain why this is terrifying – or you don’t and I need to make you understand so you never independently invent this.

And (43:59):

This is just a bad idea, ok? Virtualization belongs in data centers. Putting some poor bastard’s whole OS in a VM is a prank. It’s some Truman Show shit. It’s disassembling the coach’s car and putting it back together inside the gym. It’s not remotely worth the trouble and it probably didn’t work.

My prior comment on hydrogen mobility:

Hydrogen for mass- or space-constrained mobility (eg bikes, automobile, aircraft) faces all the known problems with storing it inside inconvenient shapes and contending with the losses from liquification. Real Engineering has a video on this aspect (Nebula and YouTube) when compared to simply using battery-electric storage.

With that out of the way, I’m skeptical as to the benefits touted on the HydroRide website. Specifically, the one about storage:

Hydrogen storage offers extended longevity, surpassing 10 years, ensuring reliability and sustainability over time.

This might be true in static conditions, but hydrogen automobiles have to vent some of the hydrogen while parked, simply to deal with the buildup of hydrogen gas, since even with excellent insulation, the liquid hydrogen will eventually get warm and evaporate into gaseous hydrogen, building up pressure. The fact is that automobiles must withstand broad environmental factors, especially temperature. And we expect bicycles to do the same: how the hydrogen tank would behave in warm climates is unclear.

There’s also not that much hydrogen in the tank. The website appears to indicate 20 grams. At 33.6 kWh/kg, the total energy in the tank would be 672 Wh, putting it at par with electric bikes of similar range and speed. Any hydrogen losses would be balanced against battery capacity loss over time.

Overall, as the article states, the target audience of rental operators might still be inclined to go with battery electric bikes rather than hydrogen. Requiring a supply of pure water in addition to electricity at charging locations – compared to just electricity for battery charging – is an extra logistical consideration. The “charge” time of 5 hours for 20 grams of hydrogen is also a potential issue.

I am not an expert bike mechanic by any means. With that said, I can’t quite visualize how re-truing the whole wheel will address a rotor rubbing issue.

The rotor mounts onto the hub at one of the most durable parts of the assembly, adjacent to the bearings. Generally speaking, the act of truing a wheel is to manipulate the rim so its axis of rotation matches the axis of the hub, where the spokes pull the rim into submission. This process shouldn’t affect the rotor, since that would suggest the hub itself is not spinning true; that could indicate an outright defective hub.

Are you able to confirm that the disc rotors are true? If the shop built and delivered the wheel with the rotor attached, presumably they checked both rim and rotor for trueness. But if you installed the rotor yourself, you might need to true the rotor.

If the rotor is true and the whole wheel is true, then that just leaves the brake pads and calipers, which could be misaligned. Although I’m not sure how this would look.

There was a ton of hairbrained theories floating around, but nobody had any definitive explanation.

Well I was new to the company and fresh out of college, so I was tasked with figuring this one out.

This checks out lol

Knowing very little about USB audio processing, but having cut my teeth in college on 8-bit 8051 processors, I knew what kind of functions tended to be slow.

I often wonder if this deep level understanding of embedded software/firmware design is still the norm in university instruction. My suspicion has been that focus moved to making use of ever-increasing SoC performance and capabilities, in the pursuit of making it Just Work™ but also proving Wirth’s Law in the process via badly optimized code.

This was an excellent read, btw.

I’m nowhere even remotely comparable to a proper furniture maker, but I can tell you some pitfalls to avoid.

Don’t cut wood without eyes, ears, and face protection. The dust, noise, or fumes will get you one day or another, if without protection. I prefer earmuffs over earplugs, but if earplugs then use the ones which tether both ends together. For a face mask, I like low-profile half-masks like this one: https://www.kleintools.com/catalog/respirators/p100-half-mask-respirator-sm

Resist the urge to dive into woodworking by starting with reclaimed wood. For example, pallets are a cheap/free source of material, but it’s a hodge-podge of different varieties, all riddled with nail holes, dents, and brown stains from rusty fasteners.

That’s not to say it can’t be done, but it certainly aggravates the process if you’re just starting. I once came across a section of 2x4 recovered from a pallet, thinking that it would cut just like the pine I was used to. Instead, it wrecked two drill bits and burned a circular saw blade as well as itself. I later mailed a sample of it to the USDA Wood Identification Public Service, who informed me that it was Acer (Hard Maple). Up until then, I didn’t even know that maple came in both varieties.

It seems hard maple is tougher than nails drill bits. I’m still learning.

Point taken. I’ve edited the title to include both measurements in inches.

It’s the curse of being engineering-minded in the USA. Much of homegym stuff here is in US customary units, but competition-inspired weightlifting equipment and anything engineering-related is in metric/SI.

Best I can do is to start with the original units (for search engine indexing), provide the conversion for people using the opposing unit system, and then try to do my work consistently within just one unit system.

Looking at this photo as an engineer, I decided to have some fun and see if there’s a way this design could somehow be a unicycle after all, based on no prior knowledge.

Supposing that maybe the two wheels turn independently, this just ends up being topologically identical to the so-called hoverboards. Except that hoverboards are configured with small diameter wheels and set to a wide gauge, allowing them to make zero-radius turns without scrubbing (aka tank steering). But this contraption, with its large contact patch and narrow gauge, would surely scrub quite hard for zero-radius turns.

Ok, so that doesn’t really work. Maybe it’s a solid axle. But… that’s somehow even worse since its performance is nearly identical to a single, wide tire. The Onewheel skateboard is proof positive that a single, wide tire is valid, but the crucial aspect is that the Onewheel still banks like a bicycle or airplane, rolling onto the edges of its tire. So too must this Inmotion unicycle, but then the outside tire is less involved in stabilizing the turn.

The only example I can think of where it’s beneficial to split one large contact point into two smaller ones is at sea, where a catamaran uses two parallel hulls for its hydrodynamic benefits. And for sharp turns, one hull lifts entirely out of the water, which is desirable. Whereas lifting a wheel off of land is not.

I can only see downsides with the dual wheel, solid axle unicycle design, reminiscent of this video by Practical Engineering on rail wheels (Nebula or YouTube). In short, optimizing for stationary stability means giving up dynamic stability while in motion, either inhibiting a turn or turning uncommanded. Trains use solid axles with conic wheels for dynamic stability and rely on truck/bogie geometry for stationary stability.

TL;DR: I can’t come up with a good reason for this design

When it comes to what insurance does or doesn’t cover, the best answer will come from the text of the policy itself. This is, unfortunately, very dry reading and most people – although instructed to keep a copy handy – don’t have the full text nearby. That said, because of the regulated nature of insurance in the USA, standardized forms of policies exist, and homeowner policies are no exception.

The common homeowner policies are numbered HO-1 to HO-8. HO-1 only pays out only for the ten listened “perils”, and is thus the most barren policy available. Not all HO-1 policies are verbatim identical, but the gist usually matches.

We can look at this sample text from a random HO-1 (issued by American Family Insurance). Page 5 shows that “fire or lightning” is covered, so that’s a good start.

On page 6, we find the exceptions to the coverage, so if any of these apply, the policy will not pay out. Nothing in Part A would seem to apply to a DIY LED project, unless you tell me your LEDs are radioactive. Part B also doesn’t apply, unless you’re somehow perpetuating a fraud using LEDs.

Part C reads like it could apply, because it mentions “construction”, “design, workmanship or specification”, and “maintenance”, but this section only applies to the dwelling and so refers to those things which are permanently affixed to the house. That would include things like ceiling fans and light fixtures, but wouldn’t include stuff that is attached to the walls using thumbtacks or 3M Command strips. It even says that:

However, we do cover any resulting loss to property described in Coverage A - Dwelling and Dwelling Extension not excluded or excepted in this policy.

This clause basically means the exceptions on Page 6 should be interpreted narrowly, not broadly.

The point is, in the entire policy, there isn’t a clause that requires listed equipment, and remember that this is the most bare bones policy commonly available. If such a requirement did exist, then building your own PC wouldn’t be possible, since the standards bodies do not test individual computer parts – except the PSU, because that plugs into the mains.

If a fire that damages the house does occur, the most probable causes would be due to: 1) an unlisted power supply or power brick feeding the ESP32 or the LEDs, or 2) no current limiting (eg a fuse) to cut out the power supply. Other failures like a shorted LED are unlikely to actually cause a house fire, and the insurance companies and UL know this; they’re more focused on preventing arc-faults that contribute to an estimated 50% of electrical house fires every year.

Good design and clean installation on your part, and using properly listed low-voltage power supplies, will mitigate the major fire risks, leaving just software bugs and lighting snafus for you to deal with.

As a matter of completeness, if there is an unlikely fire, be it from an LED project or from a candle falling over, the insurance company will still pay. But big or small, the claim will be recorded in the CLUE database along with the payout amount. This often reflects negatively on homeowners, so future rate increases may occur. But that varies by state. In any case, though, the insurance policy has still done its job: cover a non-intentional loss.

I would nevertheless advise you to have a look at what sort of homeowner policy your dwelling is covered under. Everything beyond HO-1 is nicer, and some even include limited claim forgiveness of some kind (for a price). Also consider talking to your insurance agent, who should be able to help interpret how the policy applies.

A sturdy, used, road-oriented bicycle. $200 won’t get anything too fancy, but cycling is a low-impact activity that – given the right places to bike – is meditative, improves cardio, facilitates independent exploration, and also happens to double as transportation.

I specifically say “road oriented” because I don’t want to necessarily endorse all road bikes, like the ones with carbon fibre or “Tour de France” pedigree. Likewise, mountain bikes with full-suspension sap energy away from the steady cadence ideal for a good workout, in addition to generally costing more or delivering less-than-stellar performance at low price points.