Hacker News Reader

Comment by Centmo | Hacker News Reader

I was a core engineer in a different manned eVTOL project, so have spent a lot of time thinking about this technology and watching the space. Lilium raised some eyebrows recently with their SPAC IPO valuation of over $3B, without having released footage of a single manned flight. The prototype is unique and aesthetically appealing, but my engineering mind recoils at the complexity of the design. The variable pitch blades, the adjustable exhaust nozzles, the tilt-wing vectored thrust system, etc. With complexity comes cost, maintenance, and exponentially increasing failure modes. Safety is paramount in aviation, so all flight-critical failure modes require redundant backup systems, which further increase cost and complexity. It's a slippery slope.

The eVTOL space is currently seeing a Cambrian explosion of different design concepts, but the first to see economic viability I think will be the simpler ones.

View on HN

The whole thing hinges on battery energy density more than anything else. The planning document says 20 seconds of hover at landing, with a 60 second reserve. The demo flight has 45 seconds of VTOL mode/hover at landing. That's cutting it close. You don't get a second chance at landing. No go-around.

Regard this as a bet on improved batteries. With current batteries, it's a nice demo.

This was my take as well - and was wondering what is Fig. 11 based off of? We can predict, but we don't know how the market for batteries will form in the next decade or so. I.e. the COVID disturbed all sorts of markets bad enough that price of many commodities went up X-fold and haven't gotten down yet.

Sounds basically like the Apollo lander...only today we have better computers and control systems.

The Apollo Lunar Module had the ability to abort landing by dropping the descent stage and firing the ascent engine. So if they had run out of fuel while searching for a safe landing site they still had a safe alternative option.

On the other hand, Apollo had no option than to abort to orbit. Just ditching somewhere and hoping for the best wouldn't have been an option on the Moon, but it's commonly done in aviation on Earth.

Yes, and it's a bet they will probably lose, if they really need 500 Wh/kg in three years time. Their battery graph is a fairytale at best.

The current state-of-the-art batteries that researchers build in laboratories that last 10-20 cycles can put out about 400 Wh/kg. You're simply not going to achieve first a 20% improvement over that, then a development of the tech to increase lifetime to something useable, and then scaling that to widespread commercial availability, within three years time.

If we speak about what's available in laboratories, Lithium–sulfur battery gives around 550 Wh/kg with up to 1500 charge-discharge cycles ([1]). Not available commercially, but there are a few companies that try to commercialize them ([2], [3], [4]).

I would not bet that these batteries arrive fast enough to save Lillium.

1. https://en.wikipedia.org/wiki/Lithium%E2%80%93sulfur_battery

2. https://lyten.com/

3. https://oxisenergy.com/

4. https://www.thebatt.com/news/lithium-sulfur-battery-startup-...

Yeah, I'm kind of assuming Li-S batteries won't work out. Sony announced in 2015 they were going to have commercial production in 2020. No news since then. Many others in that field have gone into bankruptcy, including one of those you link to.

Nice catch about one of the company being bankrupt.

I've taken a deeper look at Lyten and it does not seem to be a savior here either:

1. They need to produce graphene at scale and they don't yet have a solid plan to do that.

2. Lyten batteries are only 200 cycles.

Agree their chart is a fairytale based on current chemistries. It's basically a bet on commercially viable solid state batteries, which are targeting somewhere in that 500 Wh/kg range, but are probably a couple more years out than Lilium needs. I guess they could drop their viable range quite a bit at the start to compensate...

I recall Vaclav Smil saying batteries are consistently about 2-3% per year over the long haul. It's not a Moore's law scenario, but a very slow crawl. "Miracles might happen," he says, "but would I bet on a miracle happening? No."

This link is to the point in the video where he says this, but the whole talk is quite interesting. https://youtu.be/gkj_91IJVBk?t=1705

Edit: here's a more recent video of Smil talking about energy density of batteries and the problems there, specifically mentioning the unlikelihood of flying on battery anytime soon. https://www.youtube.com/watch?v=91Cs-ZkAjCg

First Law of Aviation: Take-off is optional, but landing is compulsory.

One of my senior projects in college was designing a ducted lift fan UAV. One of the things that was apparent from digging through the old research was the monstrous total complexity even of conceptually simple designs that had 1-2 ducted fans embedded in the body or wings. Computers make the complexities of controlling these things tractable, which was a much bigger problem in the 1960s. But even with relatively modern technology e.g. Lockheed has had a hell of a time with the F-35B.

Does the turbulence from the fan downwash interfere with the wings when near the ground?

yes.

When near the ground and moving slowly the wings are stalled and not providing any lift; they're essentially just dead weight. A bigger risk near the ground with older turbine engine VTOL designs was ingesting hot exhaust gasses back into the intake, thus causing a sudden loss of thrust. The F-35B mostly resolved that by using a shaft driven lift fan in the forward fuselage.

> The eVTOL space is currently seeing a Cambrian explosion of different design concepts, but the first to see economic viability I think will be the simpler ones.

Mind expanding a bit on this? And curious where you are now instead of this space... is it because you see the tech too far out?

My original interest began with the SkyCar 400 from Moller. The same design theory of a lot of eVTOLs today; just with a petrol motor. The petrol motor reduced responsiveness of the control system, but I always thought it would work out in software.

eVTOLs are wonderful, but I think any eVTOL without a 'hot swap' battery is going to be DOA in terms of the personal transport revolution.

I ended up moving on for personal and geographic reasons, nothing to do with the tech. The tech is all there for short-haul flights with batteries as the only limiting factor for range and speed.

There are various paths to economic viability for an eVTOL aircraft but I don't see how Lilium's current prototype fits into any of them. Unlike an ultralight design, it will require a certified pilot to chauffer passengers around until the day when full automation moves into the aviation space. Other than selling as an extravagant toy, this limits it to air taxi service. The design is complex so will be very expensive to buy and to maintain, so I don't see it as an improvement over existing helicopters beyond the 'green' energy source and claimed lower noise.

I am familiar with the Moller SkyCar. Lots of promises were made and broken, and lots of money burned on that project. Many modern eVTOL concepts suffer from the same basic design flaw: forward flight hanging from small propellers is inefficient. If you care about range, you need the lift of a large airfoil. Moller also had stability issues, but the responsiveness of electric propulsion and the advances in IMUs have solved this problem.

I think you can have a viable eVTOL product without a hot-swappable battery in the same way that electric cars have shown.

Their marketing has been all about taxi services for years. It's unclear if you'll ever he able to purchase one.

A landing pad for something like the Lilium would get a lot less NIMBY attention than a helicopter pad though, which is often a major factor in decision-making.

For air taxi service do you think the battery can be recharged in the time it takes to unload passengers, service the cabin, and load new passengers? If there's a long delay for charging that kind of kills the economics.

They could swap the batteries out, maybe.

I get a Magic Leap One vibe from this company. Promises, promises, promises, then reality doesn't quite meet expectations.

I don't think that Lillium uses variable pitch blades. Also, using multiple fans will lower the chance of catastrophic failure due to loss of one or more fans. For example, A system with only four fans would mean loss of a single would be truly catastrophic.

If Lillium uses multiple battery packs and redundant wiring and controls, they could achieve excellent redundancy; to a level much lower than many competing in this space.

Agree re simplicity. Ehang has been using what is essentially a consumer drone design for some time, and has demonstrated success with many flight tests. Although they aren’t using the most efficient design, I wouldn’t be surprised if they dominate the short range air taxi field.

The Ehang AAV does fly, but it seems like a commercial dead end. They aren't going to get type certification to carry paying passengers under the aviation regulations that apply in most countries. The reserve endurance is too low, and the technology for reliable autonomous operations is way too immature.

Exactly. And a seven-seater gets to pay for this.

The V-22 was how many years late? How many billions over budget?

Bell also basically had this same exact dream of tilt-rotors when they were designing and building the V-22. This was a huge part of the sales pitch of titlrotor development in general at the time.

I’m still highly skeptical of where this kind of vehicle, even if you can make it very quiet, fits in to reality. There are so many hurdles to adoption in the regulatory aspect alone that it seems intractable, but the economic aspects feel like the true nail in the coffin.

Meh.

The US is somehow singularly incapable of implementing decent public transport. Thus venture capitalists are throwing stupid amounts of money at exotic technology that could potentially alleviate some of the problems arising from this stupidity.

First we had the "full" self-driving cars and robotaxis coming any day now. Then there's Musk's attempts at tunnel digging and Hyperloop that have gone nowhere. And now eVTOLs that will go nowhere.

It's kinda fun to watch from afar though.

Don't worry, EU has money to throw around on stupid projects as well. Even worse, it's public money.

https://seenews.com/news/croatias-rimac-to-get-200-mln-euro-...

Oh sure, and countries in Asia as well, we don't hesitate to join in on these things.

I'm just saying, if you trace these ideas back to their original motivation, it's the lack of decent public transport in California specifically and the US generally.

Roads required public money to build though.

With trains you don't need self driving cars!

Rather, why don't we imagine a differently configured society? If you have streetcar suburbs, then it's possible to have a lot of last mile delivery delivered by light rail.

Besides the high-tech hype, there's also the economic absurdity behind all these proposals. The average American household already spends 1/6th of its income on (car) transportation. These VTOLs are cars on steroids, with even more modes of failure and imaginably more regulation. I assume they would do a ride-sharing / taxi business model, but even then... the average person won't be able to afford a ride in one of these because the average American can't even keep up with ordinary car payments.

It's depressingly comical to see these futuristic pitches when the reality is there won't be any customers left because the middle class will be completely gone in near time. People need to have money in order to spend it.

All air taxi services are targeted toward affluent customers, not the middle class. That's just the nature of the industry. That doesn't necessarily mean it's a bad business plan — Ferrari does pretty well targeting those customers. But even if the industry is viable that doesn't mean Lillium will succeed.

when you say "simpler", do you mean something somewhat more conventional that is a multi-rotor/tilt-rotor design like the joby evtol?

From a safety perspective (ignoring cost, though which is also critical for viability), I mean minimizing the number of 'Jesus bolts' in the design - single components that if they fail, you are in a world of trouble. In the Lilium for example, what if the tilt mechanism for one of the main wings locks up while in cruise, and you only discover this while coming in for a vertical landing? This could be due any one of many possible reasons: SW bug, electrical failure, mechanical failure, debris jamming the mechanism, etc. You've got one wing producing forward thrust and the rest producing vertical thrust which is not likely to end well. You need to either prove that every part of this particular system is extremely reliable or you need a redundant backup system in place. Due to the design, neither of these options are easy here.

Obviously you need to assume there will be motor-out situations in any eVTOL design and the system needs to handle this with full payload over the flight envelope needed for a safe landing. Lilium has the advantage of many propulsion systems to take over the lost thrust, but many other eVTOLs are questionable here.

The same thinking has to be applied throughout the entire aircraft design (electrical and mechanical) and then you need to test, test, test. Each failure mode, using full payload over the full flight envelope, stability testing under differing weather conditions, the list goes on. To give you a rough idea, having a flying prototype is great but once you are comfortable enough with the design to put in a test pilot under controlled, monitored, and ideal conditions, you are maybe 20% of the way towards something you can responsibly put into service. If you want to see a simple design, check out the Opener Blackfly.

Is an airframe parachute essentially a requirement for a vehicle like this to be viable?

It certainly increases the safety factor, but should never be relied upon as the sole redundant backup system. There are situations where it will not do you any good such as hovering at low altitude. If the aircraft does not have a reasonable glide slope, most people would feel a lot more comfortable with one.

The Lillium aircraft has no glide slope at all. It has wings, but there are no control surfaces to maintain a glide in case of power loss.

More cost.

An airframe parachute is included in the Lilium design. At least it was last time I checked.

Reply on news.ycombinator.com