How Boom uses software to accelerate hardware development

https://bscholl.substack.com/p/move-fast-and-dont-break-safety-critical

by flabber • 12 months ago

108 107 12 months ago

bscholl.substack.com

Aurornis• 12 months ago

XB-1 is the world’s first independently-developed supersonic jet, breaking the sound barrier for the first time in January, 2025. It was designed, built, and flown successfully by a team of just 50 people

This is a great headline and very impressive. However, it’s also somewhat puzzling to see the company spend so much investment money to build a small prototype plane that doesn’t resemble a commercial airliner in any way, break the sound barrier 6 times, retire it, and then conclude they’re on their way to delivering commercial supersonic passenger planes in five years

Boom Aero is one of those companies I want to see succeed, but everything I read about them tickles my vaporware senses. Snowing off a one-off prototype that doesn’t resemble the final product in any way (other than speed) is a classic sign of a company spending money to appeal to investors.

Retiring the plane after only a few flights is also a puzzling move. Wouldn’t they be making changes and collecting data as much as possible on their one prototype?

jandrese• 12 months ago

My take is that they felt like they were already pushing their luck with the prototype and didn't want to scare investors away when it inevitably crashed.

I share your skepticism, especially with their timeline. It has been some time since I looked at them closely, but they originally pitched developing their own supersonic capable turbofan to power their eventual production model. Especially with such a small team that seemed overly ambitious to me.

exabrial• 12 months ago

Hah.... in the back of my mind: announce they're going to crash it before the fly it.

"This flight we're validating our model by pushing the real world to the limit. It should explode about 38s into the test and crash. We've cleared the expected area"

_moof• 12 months ago

I work in aerospace and I don't find this development strategy unusual prima facie. I don't know if Boom is explicitly doing rapid spiral development, but this is what it would look like from the outside - a development vehicle that doesn't resemble the final vehicle design in many ways, but does have strategically selected commonality to validate and buy down risk on specific subsystems and operational concepts. They may be retiring XB-1 simply because they got the data they needed.

That being said, I share your skepticism of Boom as a company. As far as I know, they still don't have an engine for their production aircraft design.

notahacker4• 12 months ago

Yeah.

The demonstrator was to validate some basic concepts they were promoting about being able to achieve supersonic flight without supersonic booms. It achieved that at relatively low cost, and gave them something to brag about, an indication of baseline competence at certifying airframes and possibly ticked off some investor boxes. There wasn't much more to be learned about large passenger jets using their intended custom engines from a small GEJ85 powered platform, so its not surprising they haven't gone to the expense of continuing to fly it. It's not going to be useful for most other stuff they might want to test, apart from perhaps their intended custom engines which are probably years away from being certified for flight tests, never mind hitting performance and reliability targets.

rjsw• 12 months ago

The HP.115 [1] and BAC 221 [2] were not exact scale replicas of Concorde.

[1] https://en.wikipedia.org/wiki/Handley_Page_HP.115 [2] https://en.wikipedia.org/wiki/Fairey_Delta_2#BAC_221

sidewndr465• 12 months ago

The market for Boom is not commercial passenger flights. So much time is wasted with security, boarding, taxi-ing, waiting at the destination for a gate to unload, etc. that the flight speed is not a big deal. Existing commercial passenger jets could already go faster without going supersonic and save some time, but it doesn't matter. Even if you fly commercial passenger jets at the absolutely face-melting Mach 3.3 of the SR-71, you don't really save enough time to matter. The maximum speed in flight doesn't do anything to address ground delays.

JumpCrisscross7• 12 months ago

time is wasted with security, boarding, taxi-ing, waiting at the destination for a gate to unload, etc.

Airlines can optimise for this. Digital ID virtually eliminates security lines. Paying up for gate, t/o and landing spots takes care of the latter. There is a cost tradeoff for service in the airline business. An all-business airline flying Booms would almost necessarily have to pay up to negate these issues. (That or fly out of the FBO terminal.)

SkyMarshal2• 12 months ago

That may be true for domestic coast-to-coast flights, but not for transoceanic ones across the Atlantic, or especially the Pacific, or north-south across hemispheres, that can take 8+ hours. Flight time is a higher portion of the total travel time in those cases, and seems like the main market for Boom, especially if they initially target Business Class flyers who do those kinds of trips regularly.

testing22321• 12 months ago

Not disagreeing with you at all.

What is the market for Boom?

cyberax• 12 months ago

I see you never flew from LAX to ICN.

SkyMarshal2• 12 months ago

I think part of it was that they were testing a new aerodynamic design that eliminates or minimizes sonic boom, so they can go supersonic over land almost immediately after takeoff, and operate over populated land routes. It makes sense to test that kind of thing with the smallest possible model first, then see if you can scale it up to passenger size without losing that quiet acceleration. Their timeline for doing that may be optimistic, but what they're doing makes sense.

dingaling2• 12 months ago

The XB-1 doesn't have any boom reduction shaping. That's the NASA X-59, though that aircraft is pretty much a dead-end in that it's not scalable to a passenger configuration.

The XB-1 made use of an atmospheric trick to minimise boom propagation to ground level on one test flight, so well-known in fact that Concorde sometimes used it to accelerate as it coasted-out without an audible ground-level boom. Unfortunately that trick runs out at about M1.17.

dingaling2• 12 months ago

It's also largely PR guff. The first privately-developed supersonic aircraft was the Northrop N-156F, forerunner of the F-5, that first flew in 1959. Funded entirely from company funds with no military contract. And it went supersonic in its first flight with no drama.

In fact the chase plane for the Boom XB-1 is a T-38, derived from the N-156F. It can outrun the XB-1.

bangaladore• 12 months ago

I'm not sure how strictly privately developed the N-156F is given you could easily argue that reuse of design, knowledge and relationships from existing contracts saved them a lot of money.

everfrustrated2• 12 months ago

Their immediate goal is to get the next round of funding. Viewed from this lense it makes a little more sense.

Twirrim• 12 months ago

Together with a few other optimizations, these tweaks yielded over 1,000mi in increased range—enough that we could now afford a remarkable passenger cabin without sacrificing fuel efficiency or range.

Honestly, the way the narrative reads, they're still sacrificing 1,000mi of range in the interests of an improved cabin experience. They've just found an optimisation that enables them to reach a net neutral state.

Given we're effectively talking about fuel efficiency here, it's hard to imagine airlines wanting an improved cabin vs less fuel consumption. All the incentives are on them already to meet a "barest minimum" cabin experience that they can get away with, because every bit of luxury costs them in numbers of passengers, and fuel costs.

wayne• 12 months ago

It might be a fad, but the current trend in US public aviation is increasing premium cabins and premium revenue: https://simpleflying.com/why-us-carriers-doubling-down-premi...

This is the reason Delta and United and doing well right now and Southwest and the LCCs are struggling.

It wasn't true just a few years ago, but if this continues as a trend, I could see an airline sacrificing fuel efficiency for a dramatically improved onboard experience.

notahacker4• 12 months ago

Premium cabins tend to be a very small proportion of overall seats and are about overcharging for a little extra legroom and service rather than trading off operational flexibility for unique luxury though. Big difference between charging 3x economy rates for 2x the space for a carefully estimated proportion of seats in a mixed configuration (no brainer) and hoping your layout is so good it justifies thirstier, less flexible aircraft to operators (tough sell)...

That said, Boom's customers - if they ever exist - will be a new business class pay extra for supersonic flights category anyway.

sandworm1013• 12 months ago

It isnt a trend. This is marketing. Thirty years ago, the a380 was pitched as having room for luxury too. The new plane is always going to have more legroom, wider aisles and better air conditioning than anything before. But it never happens. The pitch to actual operators is the square-feet of floorspace and how many seats can be crammed into that space at given price points. Just like concord, this thing only makes sense with quazi-economy seating. Do not expect to nap on a nice lie-flat seat.

cogogo• 12 months ago

Saw a Jet Blue plane wrapped in Peacock livery today… selling the planes themselves as billboards sure does feel like scraping the bottom of the revenue barrel.

handfuloflight• 12 months ago

That would be an accurate statement if they were being sold for bottom of the barrel prices.

deepsun• 12 months ago

To be fair, modern airliners, even budget ones, are way more comfortable than Concorde. You can visit one in a museum, it's very cramped, and noisier. Concorde had way better service tho.

highfrequency• 12 months ago

We can literally define an airplane parametrically in a configuration file and press a button. In a matter of minutes we have a complete quick-and-dirty analysis of how the whole aircraft performs—as mkBoom flies the aircraft through a full simulated mission (takeoff, climbout, acceleration, cruise, descent, landing). Overnight, mkBoom can run higher-fidelity simulations for a more exact understanding of performance.

Awesome stuff! Allows large scale exploration across all dimensions of plane design to jointly optimize all components and their interactions.

polishdude20• 12 months ago

I wonder if they run this through an optimizer then?

signatoremo2• 12 months ago

Boom’s potential customers wouldn’t be able to put more seats on their planes even if they want to. I suppose the targeted performance affords very little margin for customization

dreamcompiler4• 12 months ago

So they can script their design software then. In the modern era I'd consider this table stakes for any big engineering project.

theptip• 12 months ago

As an aside, anyone care to speculate on the “secret seat configuration”?

I guess maybe it’s a recliner with feet pointing to the outside (maybe just two seats per row)? That’s the only new configuration I can imagine that would require reshaping the hull.

Etheryte• 12 months ago

I'll be very disappointed if the big secret isn't the smart-fella-fart-smella configuration.

sandworm1013• 12 months ago

Forget luxury. Forget speed. Hands down, the best flying experience I've ever had was on a dirty, slow, late and loud C-130. After an announced delay on the ground, I wedged myself between a cargo pallet and the wall, threw a ratchet strap across as a "belt" and passed out on a metal floor with a backpack for a pillow. No in-flight meals. No safety briefs. No entertainment systems. No drink service. Nothing. I don't even remember the takeoff. The only thing anyone said to me was "Uh, sir... We are about to land. You have to get up." THAT is what I want from flying. Give me a bit of peace, a chance to sleep, and I couldn't care less how fast or slow the journey.

JumpCrisscross7• 12 months ago

You’re describing a lay-flat seat. (If you let them know you don’t want to be disturbed, they won’t.)

vosper• 12 months ago

Of if you are on Air New Zealand and can't afford business class you can get a Skycouch in Economy. They're pretty great, actually, unless you're over 6ft tall or can't sleep with you knees bent a bit

sandworm1013• 12 months ago

Are you kidding? On a commercial flight, between the safety briefs, seatbelt warnings, and "turn on/off your devices" there are constant announcements. And the stupid entertainment systems you cannot turn off, or at least that spring to flashing life again after each pointless announcement. I wore my ear defenders on united once, and was woken mid-flight by a steward informing me they were not allowed as i wouldnt be able to hear announcements.

dgunay• 12 months ago

I know it's due to safety but I really wish there were an economy version of the lie flat seat. I'd gladly sleep in a set of bunk beds stacked 3-4 high for an 8 hour flight.

TheHideout3• 12 months ago

AKA they're doing what every other aerospace company has been doing for decades, multidisciplinary design analysis and optimization [0] with simulation in the loop. If you were to ask them how they're leveraging Design of Experiments I bet it'd be met with "design of what?".

[0] https://en.wikipedia.org/wiki/Multidisciplinary_design_optim...

class3shock4• 12 months ago

That doesn't sound good to investors that don't know anything and don't do any due diligence.

fluorinerocket2• 12 months ago

Shhhh we're calling it AI now

saagarjha• 12 months ago

I’m a little confused how this differs from standard constraint optimization.

TheHideout3• 12 months ago

In standard constraint optimization you know the constraints at compile time. In MDO many constraints are generated at runtime and constantly change as you search for solutions.

class3shock4• 11 months ago

It doesn't, they just want to use a buzzword to make it look like they are doing new and advanced things.

ricardobeat2• 12 months ago

Which non-military aircraft have been developed with these methods in the past 20 years?

pxeger1• 12 months ago

Boeing 787

klysm• 11 months ago

On what basis do you think nobody on the team understands DoE?

throwaway098909• 12 months ago

Looks like investors and potential employees are asking if Boom is fast growth tech company and this seems like a lame attempt to make up an answer instead of saying- we're solving a different problem on a different timeline. Can't imagine Elon writing something like this 10 years ago.

deterministic• 12 months ago

The key take away is to embed software developers everywhere in the development process.

Which makes perfect sense. Software is about automating things. And the more you automate, the faster you go.

A bit like how Jane Street operates I think.

class3shock4• 12 months ago

TLDR: We changed our plane design and lost range, we used ai in helping tweak our engine design to get back our range, we were successful in our computer models.

This article is just another from what are fundamentally hardware companies proclaiming how they're implementing "ai" and it's doing so much to help to gather more stock/venture investors but with no actual substance.

Everything he discusses taking into account are things that airplane/engine manufacturers already do and it doesn't require ai, just some python or god forbid, an excel sheet.

His statement about the relationship between airplane and engine manufacturers is factual correct but so blatantly wrong. You think GE/PW and Boeing/Airbus engineering don't work closely with each other when a new model is being developed? You think either would risk hundreds of millions and years of development without talking to each other? How does anyone take this guy seriously enough to give him millions of dollars?

"This close approach gave rise to mkBoom, our proprietary airplane design software. Initially created in a simpler form for XB-1, mkBoom has evolved significantly and is now pivotal to designing our Overture airliner."

This is the same overture airframe that supposedly had it's final design released in July of 2022? And was supposed to be rolled out in 2025? And have it's first commercial passengers in 2029? (Ref. 1)

But apparently 18 months ago the structure was being redesigned in such a way that caused a 1000 mile range loss (that is described as "subtle" fuselage change)? And in response to that last year they completely redesigned the engines and regained that 1000 miles of lost range... Sure.

More importantly, given the FAA states they a new aircraft takes 5-9 years to certify (Ref 2.) it seems the 2029 target is not viable any longer right? So are you going to tell American or should I?

Ref. 1 https://news.aa.com/news/news-details/2022/American-Airlines...

Ref. 2 https://www.faa.gov/aircraft/air_cert/airworthiness_certific...

fbd_01002• 11 months ago

Agree on most points. However as someone who has worked on new aircraft I can tell you that friction/mis-communication/finger-pointing between airframe and powerplant manufacturer is a huge issue. There is a common saying in the industry, that you should never design a new airframe and a new powerplant at the same time. Each new aircraft certification program should be one or the other

class3shock4• 11 months ago

Sure, as there is when any two companies work together but to say they have no ability to communicate performance parameters, that they have no ability to work together, is just wrong.

dreamcompiler4• 12 months ago

At Boom, every engineer is expected to code and to leverage AI.

As an engineer I find "leveraging AI" to be a very troubling idea. I'd want to know detailed specifics of just what management believes AI should be used for before I accepted such a job.

As a passenger, I damn well don't want to fly on an airplane designed with software that can't count the number of bs in blueberry.

johnvanommen• 12 months ago

I listened to the podcast that the founder did about a week ago. It reminded me of how retired folks in the middle class open a restaurant when they don't know anything about running a restaurant. Except this dude isn't investing $1M on a McDonalds, he's investing hundreds of millions.

He seemed almost proud of his inexperience, and nearly said that it gave him an "advantage" because existing engineers weren't willing to "innovate."

ricardobeat2• 12 months ago

This is a common story in the startup world. Outsiders are able to break free from the mould; its harder to innovate when everything you do is already shaped by best practices, and your career is highly dependent on your peers’ approval. Doing things differently is a high risk move that few are willing to make.

Incidentally, Ray Kroc, the guy who made McDonalds the $200B company it is today, didn’t know anything about running a restaurant. His closest experience was selling blenders.

avmich• 11 months ago

SpaceX did Falcon-1 a few times cheaper than assumed, and that was the first aerospace experience for Elon. Turned out impressively good...

dreamcompiler4• 12 months ago

They keep calling this thing an "airliner." Assuming they ever go into production, they'll definitely sell some units to Saudi princes, but commercial airlines? Nope. Any customer that could afford a seat on this thing could afford to rent a Gulfstream and not have to share it with others.

muragekibicho• 12 months ago

I always thought his experience at Groupon made him a master of Reed Solomon error correction.

I saw Reed Solomon codes were invented for fast, accurate missile guidance and (in my head) that's the connection I made to his super sonic startup

tinfever• 12 months ago

Anyone know what their simulation stack might be using? From the modest amount I know, mechanical engineering usually dislike their CAD tool and Ansys FEA simulations don't always converge. So how on earth are they simulating an entire aircraft with a single command?

When they say simulate, so they actually mean just using ideal mathematical models in Matlab?

fbd_01002• 11 months ago

I work in this space. Most likely it is a huge conglomeration of commercial CFD, FEA, and CAD softwares, custom in-house codes, and empirical aircraft design curves all stitched together with some python.

torginus• 12 months ago

It's weird how analog the entire field of aerospace has been - I remember reading articles in the 2000s (and before), of how computational simulations a are going to unlock crazy new never before seen designs, controllable via novel control surfaces only possible thanks to novel control surfaces whose position is determined via literal supercomputers onboard the plane. And how X-planes are going to become unnecessary, because everything a wind tunnel test could tell you can be predicted via simulation.

Fast forward to today, there's been no aerodynamically novel aircraft developed in the past decades, and from what I read, wind tunnel and glider tests are still necessary to validate aerodynamics during complex conditions, like manuevering.

goodcanadian• 12 months ago

Computational simulations have been involved in aircraft design for decades. However, at some point you still have to verify the simulations with real world tests. I think (as a non-expert) the main reason no aerodynamically novel aircraft have been developed lately is because we have essentially optimised designs, already. To be clear, I am not saying that we can't do better. I am saying that the manufacturing (and other) cost of incremental improvements is large and the benefit is small. For commercial aircraft (as an example only), it is really hard to beat a tube with wings for the lowest cost of moving large numbers of people or volumes of cargo. Military aircraft with specialised roles are a bit more varied, and a lot more expensive to build and fly.

XorNot3• 12 months ago

It happens every now and again on here: someone comes up with like a 2% improvement in aerodynamics, and people are unimpressed. Meanwhile airlines are basically scrambling to get it rolled into their next-gen purchases because it's the biggest improvement in costs in a decade.

rkomorn• 12 months ago

I think there's also a lot of "if we don't change too much, we have a pretty good chance to not face issues getting the aircraft certified", and "if we change the aircraft too much it'll cause issues where airports have to alter their infrastructure" (which is relevant for things like the blended wing research that's popped up recently).

There's a ton of legacy in overall airline/aircraft operations that discourages big changes.

SideburnsOfDoom4• 12 months ago

What is Boom's current plan to make the engines that they need? How is that going?

IIRC, the last news was that Rolls-Royce noped out of that (1), which is an indicator that it's either not technically or commercially feasible.

Are they still planning to design the engines in-house?

I'm not an expert, but this seems like it's on the critical path to success, and also high chance of failure.

i.e. Without engines, they have nothing, and if Rolls-Royce can't do it, then who can?

1) https://www.space.com/boom-supersonic-rolls-royce-engine-spl...

jdeibele• 11 months ago

From the first paragraph:

"It was designed, built, and flown successfully by a team of just 50 people—compared to the hundreds or even thousands that would have been employed by a traditional big aerospace company. And we did this with roughly a tenth of the budget that would traditionally be required. People have marveled at how our small team of just 50 people at Boom designed, built, and successfully flew the XB-1, the world’s first independently developed supersonic jet. And we did this with about a tenth the capital as any other supersonic program."

Is the repetition a sign of AI writing the article?

How Boom uses software to accelerate hardware development

Related Stories