Hacker Read top | best | new | newcomments | leaders | about | bookmarklet login
The inventor of the black box was told to drop the idea (spectrum.ieee.org) similar stories update story
216.0 points by sohkamyung | karma 76115 | avg karma 9.95 2021-06-19 02:05:10+00:00 | hide | past | favorite | 114 comments



view as:

Well if this idea can't survive being dropped, it's not much good is it? Sorry...

> Interestingly, modern flight data recorders are required to have an underwater locator beacon that pulses once per second at 37.5 kilohertz for 30 days and is supposed to be detectable at depths of at least 14,000 feet (4,267 meters). But they don’t have comparable beacons for crashes on land.

After Japan Airlines Flight 123, that is shocking to me.


>But they don’t have comparable beacons for crashes on land.

Well, they do: https://en.wikipedia.org/wiki/Emergency_position-indicating_...

But the reason you need underwater beacons in particular is that the water absorbs RF signals, so you'll actually use acoustic beacons underwater. Such a device would be useless above the water since air is so much less dense.


From what I understand in America, small airplanes are required to have ELTs/EPIRBs, but airliners are exempt. I don't know why that would be, maybe lobbying.

Possibly their size means they're remote trackable (e.g. radar) already by gov/military when over or near land. This may not be the case with small aircraft.

Total guess, of course.


The black box has worked well for all in Common Law countries.

In countries with French-style law, that might be different for pilots. (Essentially you're guilty until proven innocent, and they hire a retired judge with nothing better to do except to drag out the hearings (and fees) until they die.)

However the scale has tipped recently in the US with a central pilot records database being created recently. So at every step of your career, you're recorded and tracked. If an employer doesn't like you, they can scribble whatever they want in your permanent government record now.

So before a pilot candidate spends $200,000 on a college education and pilot training, they need to think about both losing their medical at any time, and what will happen if their airline decides to fire them and give them a hard time.


> $200,000 on a college education and pilot training, they need to think about both losing their medical at any time

In my country Loss of License insurance removes that as a risk for pilots.


In the US there is loss of medical insurance (essentially disability insurance) available in most scenarios from several providers, provided somebody pays for it. Note there's always policy fine-print and limits.

But there's no loss of certificate insurance product available. The FAA pads any complaint with "reckless", which is up to their sole opinion.


In Roman law you're also innocent until proven guilty - and the unwritten law in most countries is that prosecution doesn't get to access accident investigation materials and has to order their own experts every time.

This inventor is probably joined by many others before him that had the same idea, were shrugged off and did not have the tenacity to follow up despite that. Another example for what is common wisdom: Execution is everything.

The myth of the inventor-founder is something from the industrual age. Interesting how it still persists.


Why is it a myth? Whenever I read about the founding of a big name company there's very often a significant innovation that it's initially premised on.

Disney - animated feature length film with sound, dogged vision of founder

Post / Kelloggs - invention of breakfast cereal, invented by founder (Kellogg's invention stolen by Post)

Carnegie steel - bridge prototype, highly risky and innovative endeavor

Ford - division of labor

Mars - caramel and nougat in chocolate, innovation over Hersheys' plain chocolate, invented by founder

Google - pagerank, invented by founders


Post and Kellog have joint tenancy in weird theories about breakfast cereal and masturbation. I wouldn't class Post as a thief in this space. Also grape nuts are fantastic (and that isn't a joke about nutting)

Mass production and factories had already been a thing for almost a century when Ford started his factories. It is like saying that the new innovation "computers" was behind Facebooks success.

No, the Ford's innovation is division of labor while producing a complex thing, a car. Instead of a team working on various operations, Ford's system introduced workers who each did a very simple operation, without having to change or readjust. The need for coordination between workers was largely removed, replaced by the conveyor, seriously improving production speed.

Division of labor while producing a complex thing was a concept well established and documented during the industrial revolution before Ford. E.g. Adam Smith had a lot to say exactly about "Instead of a team working on various operations, [...] introduced workers who each did a very simple operation, without having to change or readjust.", it's misleading to call it "Ford's system" because Ford applied it extensively, but he did not invent it and wasn't the first to use it. Conveyor lines also wasn't an invention of Ford - he applied them a lot on scale, but they were widely used in different industries long before he was born. He also wasn't the first to apply continuous assembly lines specifically to the production of cars, as far as I understand the founder of Oldsmobile had been doing high-volume series-production of cars with strong division of labor a couple years earlier, though eventually not as successful as Ford.

He can and should be credited with applying these innovations at scale and ironing out various kinks in the process and popularizing these concepts, but it would be a lie to call it "Ford's innovation", all these things you mention were initially invented by others, although Ford did manage the practical implementation of all these innovations in combination unusually well.


No, that was pioneered by Ransom E. Olds. That doesn't stop Ford from repeating this at every opportunity but it isn't really true.

https://en.wikipedia.org/wiki/Ransom_E._Olds


Which car company was using division of labor before Ford?

Olds-mobile, of course.

More recent examples are:

- Robinson Helicopters, the largest mfg. in the world by unit count

- Dean Kamen, several medical devices plus the Segway

- an endless number of Internet companies.

Note that the Google founders couldn't write their own production-level software, so hired a python contractor.

Also for those that aware, Tesla's patents were used to start Westinghouse.


This is not a myth, but History recording real facts that happened.

Anyone can imagine a "flying car" or a "nuclear fusion reactor". It is a completely different thing having and putting the resources to work: the people, the effort, the money and the time to make something real.

Someone just "having an idea" only works in science fiction, and even there you need to do the work of writing the book, or drawing the comic.


"Execution is everything" is always true after the fact. Good and bad ideas are treated with disdain because that is how people are. Until a random person gets profit from execution, then is praised for the profit, because people desire and envy profit.

If there is a myth here, it is that it matters who executed it first. It's random.


Wow. History hasn't been kind. As is often the case around limited-mindset thinking.

From TFA: "civil authorities said that “Dr Warren’s instrument has little immediate direct use in civil aircraft.” The Royal Australian Air Force likewise decided that “such a device is not required—the recorder would yield more expletives than explanations.”

Well. No. Multiple air crashes across the world relied on recordings as a key part of their investigation. Especially when a plane lands in the Hudson, just to name one example.

This is very telling:

"Most damning was the Federation of Air Pilots, which declared that the device would be like “a spy flying alongside—no plane would take off in Australia with Big Brother listening.”

History wasn't kind here either. Just try and take off in a jet from a commercial airliner without the various safety equipment installed. The insurance companies alone will try to eat you alive, let alone every other agency.

There's definitely a repeating theme around safety equipment and authority/authority figures. The ridicule at the start is definitely a common starting point. Seat belts and air bags were both ridiculous inventions. Then they weren't. Then they became required.

Quote from elsewhere and not necessarily about flight recorders: "It has been said that any new idea must pass through three stages. First, it is ridiculed; second, it is subject to argument: third, it is accepted. The safety idea has reached the final stage. It is accepted." -- Earl B. Morgan, journal of Safety Engineering, 1917


Fourth: in retrospect, it is seen as obvious :)

I guess one possible explanation is that having a black box in your plane is not useful to you, only for other people after you die. Pushing for the idea requires accepting you are going to screw up and some people will think that your job is preventing the airplane from crashing in the first place. It's obviously useful in the long term from a societal perspective, not from a short-term and individual point of view.

Even then I don’t get why someone would object. But of course hindsight is 20/20 so that may be why. Having a black box in all aircraft would help me too, because other crashes are better understood. Then again I also do want my death to be maximally useful to others should I die, so that at least I don’t die completely in vain. I don’t want to live forever, but I don’t want my life to end abruptly and outside of my control either, and I have plenty of things I want to do still before I leave. So that may also be why I don’t understand that POV you suggested.

Also, if I was a pilot, I would appreciate the fact that the black box can potentially help my friends and family get closure, because if the black box survives then my friends and family can learn what happened that caused the crash.


What they might learn is that the crash was your fault. It isn't clear that's the legacy you'd really want to have.

And they might learn that it wasn't, thus clearing your name.

In the aviation world, people aren't at fault. Processes are.


I think pilot must have known that the information in the back box would provide them heaps of data on how to prevent crashes, to their benefits. My guess is that is was not opposed by average pilots, but by people in power above or around or among pilots, whose power would be undercut by any new idea.

First They Ignore You, Then They Laugh at You, Then They Attack You, Then You Win.

"The statement evolved from a large family of sayings that originated in the nineteenth century. In 1918 a closely similar remark emerged in a speech by Nicholas Klein, a union representative. Gandhi discussed stages that a movement passes through in a collection of writings he published in 1921, but his words did not really match the target expression."

https://quoteinvestigator.com/2017/08/13/stages/


"Don't worry about people stealing your ideas. If your ideas are any good, you'll have to ram them down people's throats."

Howard Aiken


Are these the 'trust the experts' everyone was talking about during Covid?

What about a black box for your car?

Yes. You can find the same kind of bs regarding safety belts in cars, ABS, airbags and so on. The part where I'm seeing a difference is when safety gear starts to actively interfere with operating the vehicle.

It’s kind of funny to see basically the same arguments being offered against video recording in this very thread on HN.

Good thing this idea didn’t go down in flames. (Sorry).

Looking back, almost every great idea I’ve had has received negative support from people around me, who either didn’t understand it or had their own plans for me.



I wonder why there was such a difference in responses between the Australians and the British. Was it just a culture difference?

But it was an interesting turnaround by the Aussies that they were the first country to mandate the flight data recorders.

Now that the technology is so inexpensive, we need to have them in cars!

Get a dashcam.

I wonder how soon a car would be able to publish its metrics, like speed, motor rpm, data from ABS and brakes, etc via a protected Bluetooth channel, so that a dashcam or a phone could keep recording it, like the black box.


Now?

You can already get those metrics from the car's OBD2 port.

There are even dongles which have Bluetooth so you can see/record the metrics on your phone, so it shouldn't take much to get them on a dashcam.


Dashcams pointing outwards are not legal in the EU due to privacy laws.

Just don't IoT it, please.

To late

Already a thing and quite widespread deployment. Mandated in the US, Japan, and Korea for many years now: https://en.wikipedia.org/wiki/Event_data_recorder

Well, ok. First you make it a crime for any authority to look at it for any reason that is not blameless safety investigations.

The last sentences about uploading data to the cloud had not struck me before. It is such an obvious development, if not to replace so to complement. But perhaps there are challenges around it that are not obvious?

A plane more often has access to an atmospheric cloud than to a computing cloud, especially when flying over an ocean. Maybe Starlink could offer reasonable connectivity, even if intermittent.


It is such an obvious development, if not to replace so to complement. But perhaps there are challenges around it that are not obvious?

You might be surprised how scant and expensive bandwidth becomes once you are out of sight of land


Still, there are so many flights now that offer onboard Wifi, it’s seems silly not to back up flight recorder data to the cloud wherever possible.

Inventor: this is my indestructible black box for recording aircraft crashes.

Reviewer: drop it.

Inventor: this is a safety feature that will allow us to iterate plane design and save thousands of lives in the future!

Reviewer: drop it.

Inventor: this is extremely important for the future of aviation!

Reviewer: drop it.

Inventor: I WILL NOT storms off

Reviewer: I just meant...show that it won't break? Never mind.


I'm not surprised.

The technically savvy HN readership is overwhelmingly against black-box designs.


I've watched all the episodes of Aviation Disasters. The device missing from cockpits is the video recorder. Many investigations revolve around questions like what is the pilot looking at, what the instruments show, even who the heck is in the pilot's seat. Video can confirm switch settings and instrument readouts (the voice recorder is also used to confirm if the warning horns sounded properly).

Instead, a lot of effort goes into trying to reconstruct this.

Just put a dang video camera in the cockpit.

I've also, for decades, advocated a camera that records flight operations at an airport. Just stick a couple in the tower pointed at the runways. Note that in the Concorde disaster, there was no video of the take-off. They had to rely on eyewitness testimony, which is notoriously unreliable. The only video came later from some passenger in a car who happened to have a video cam at hand. That accidental video proved valuable. (It's a horrifying video.)

I was surprised at all the vitriol opposition to pointing a camera at the runways. People confidently told me it would cost millions of dollars, and was completely infeasible. Jeez, anyone could buy a security camera that recorded on a loop for a few hundred. Every convenience store had them.


Couple of issues here; the first and foremost - recording people at their workplace. We can all argue how privacy is unreasonable in public but recording let alone pilots but anyone continuously at work will receive big pushback.

Also video occupies much more data than most other streams of information stored on blackbox. This coupled with the fact that anything aviation related takes decades to get into widespread use and it was only quite recently that solid state storage became so dense and ubiquitous.

Video surveillance at airports is getting more widespread, just search for SFO asiana.


It's the same issue with the audio recording. The solution is it's in a loop, and those recordings are almost never released.

As for video, VHS quality is about 1Gb per hour. Meaning you can put 64 hours on an ordinary USB stick about the size of my thumbnail. Furthermore, video recording has nothing to do with flying the airplane, so there are no safety/certification issues. It doesn't even have to be in a crashproof box.

I know that airports now have video surveillance, but those cameras are pointed mostly at the tarmac for security purposes, not at the runways for flight ops.


BTW, I can buy a dashcam for $100 that records HD in a loop and plugs into a cigarette lighter. I'm not buying there's any technical or cost issue with installing one in the cockpit. I will buy that the problems are all bureaucratic in nature.

Just make a secure bracket for it, and a 12V feed.


If you have that to add, I am pretty sure you don't have any formal experience with aviation.

Even the nuts and bolts that go on aircraft can cost >$100. The problems are not all bureaucratic, and I get HN's hate of bureaucracy but for life-and-death matters - I'd rather default to caution than move fast and break necks attitude of SV/Tech.

Just to give you an example, lookup the swissair in-flight entertainment fire incident. Tacking couple of screens and media server onto an existing plane should be fine right? turns out not really and 215 people paid the price of that decision with their lives.

Privacy aside, being watched does affect decision making.

If you think recording people at their jobs will make things safer, why not apply it for the least safe mode of transportation first - why doesn't the gov/NHTSA mandate a video and data recorder in your car and the police and insurance can take a look at it next time you crash to decide who/what caused it.


No need for video recording on general aviation, but a public transport is a completely different thing.

Any time I enter a bus in my town, I see that it has a number of cameras, which point to the road, to the passengers, and to the driver as well. Why should a plane be different?


> If you have that to add, I am pretty sure you don't have any formal experience with aviation

Walter Bright is an aerospace engineer, programmer of flight critical systems of airliners, son of Charles Bright. You are way off on your accusation.

citation: https://www.amazon.ca/Jet-Makers-Charles-Bright-ebook/dp/B00...


>why doesn't the gov/NHTSA mandate a video and data recorder in your car

The automakers were already "voluntarily" putting EDRs in cars, so no need to mandate it.

From 15 years ago, standards were set so that data can be retrieved after an accident and such:

https://www.nhtsa.gov/sites/nhtsa.gov/files/fmvss/EDRFinalRu...


> Just to give you an example, lookup the swissair in-flight entertainment fire incident. Tacking couple of screens and media server onto an existing plane should be fine right? turns out not really and 215 people paid the price of that decision with their lives.

I'm not super familiar with this one, but based on a quick skim of the TSB report (http://www.tsb.gc.ca/eng/rapports-reports/aviation/1998/a98h...) it looks like the issue here was not the IFE itself but a poor installation that significantly differed from the certified configuration and resulted in chafing on power wires.

It's also worth noting that this happened in the '90s when networked personal IFE systems were still in their infancy. It's 2021, there are no unsolved problems in installing a few cameras in an aircraft. A lot of them already have cameras accessible through the IFE, putting one more in the cockpit and a hardened DVR in the back isn't exactly rocket science.

Yes, aviation stuff often has a few zeroes added on to the price tag for good and bad reasons, but we're basically talking about a dashcam here. Low triple digit prices for non-aviation hardware, so even if you multiply the price times 100 it's still not really significant compared to the cost of an airliner.

There are no good technical reasons to not have video.


"If the people in the cockpit aren't doing anything wrong, they have nothing to worry about".

A reasonable argument for stoners lighting up in the supermarket carpark.

Not so for someone flying several tons of metal that will absolutely fuck up the lives of many people if they aren't paying full attention.


why we cannot just delete record when airplane take off next time? if airplane landing safely, we do not need record anymore.

> recording let alone pilots but anyone continuously at work will receive big pushback

I think we can say that when you have hundreds of people's lives in your hands you are accepting a higher level of oversight and invasion of privacy.

The same reason some people in some careers have to go through deeply invasive background checks.


There isn't an inch in Taipei that doesn't have a CCTV on it.

In the US I see cameras at every gas station, donut shop, drug store, liquor store, and lab I go into. None of these places are as high stakes or high security as an airport.


We record bus drivers, for goodness' sake. We should be able to record pilots.

I agree that a hardened process relying on multiple SSD (that are then uploaded and wiped after landing at an airport) would be a good technical direction.


No, if you're a pilot in charge of a few hundred lives + the lives of a few hundred more in case you crash on the ground your 'privacy at work' is trumped by the responsibilities you have taken on. That's not a good argument. What is a good argument is that people might behave different if they believed that there would be consequences for their actions, and this might either be a negative (if they don't act for fear of repercussions) or a positive (if they don't perform acts that could endanger the aircraft knowing that these are recorded).

This already happens. At least here in Brazil, offices have security cameras in pretty much every environment, except the restrooms.

An office is a shared environment, usually not owned by the people who work there, that exists only for job related activities. Just like companies can monitor your company issued email, network traffic and cell phone, they can monitor your activities in the workspace.

I'm a privacy advocate, I avoid Google and Facebook as much as I can, but even I understand the difference between private and professional life. One has an assumption of privacy, the other doesn't.


Note that this is not the case in all jurisdictions. In many European countries, the company cannot monitor employees' corporate emails without specific justification.

> Couple of issues here; the first and foremost - recording people at their workplace. We can all argue how privacy is unreasonable in public but recording let alone pilots but anyone continuously at work will receive big pushback.

Casinos do it. Banks do it. Convenience stores do it. The TSA records it's workers and you when you go through security. This doesn't seem like much of a hurdle.

For example: https://www.jwincorporated.com/projects/ohare-international-...


You’re arguing against monitoring equipment in the cockpit in an article that’s about people arguing against monitoring equipment in the cockpit. Thanks for making me laugh today. I appreciate you.

Although I don't necessarily disagree with your arguments, using security cameras on the tower would probably give you very little : the distances are huge.

Imagine the footage of the average shop robbery, which most of the time is already useless. And UFOs.


It'll tell you the positions, directions, and speeds of the traffic on the runways, which are a big deal.

Don't they have ground radar for this?

Maybe they do now, at some airports. Radar is far, far more expensive than a dashcam set in the tower.

If I was a pilot I'd hate to be constantly surveilled by a camera, even if the footage is "never reviewed". In fact, it's got nothing to do with being a pilot.

Don't buy a Tesla then. They have cameras pointing at you and recording while driving.

So do taxis, and any other car with driver assistance (though those don't save the video, they just watch if your hands are on the wheel.)

> Don't buy a Tesla then

Way ahead of you on that one.


That’s opt-in, at least in the US, and for the moment.

(Just curious) what features become unavailable when you opt out? Or in other words, why would anyone opt in?

I didn’t look too closely, since I just wanted to make sure it was off, but iirc they said they only upload the footage around the time of a crash to see how effective the safety systems were, and to get a sense of what to improve”. Basically one of those “please waive your privacy in the name of science” types of deals. Didn’t dig through any privacy policies, though, I’m sure Googling will yield a better answer.

I'm not gonna?

If I was a private pilot, sure (and I am one). If I am a commercial pilot ferrying hundreds of passengers or millions of dollars worth of cargo, I'm no different from a public transit employee (or a public employee for that matter), in which case I damn well have to put up with being surveilled on camera.

I can see the privacy issue, but your voice is already recorded. A camera wouldn't infringe that much more on your privacy...

Haha. Even public transport has had cameras for years.

>even if the footage is "never reviewed"

Why the scare quotes? It would be literally never reviewed unless incident occurs, as it would be on overwriting loop.


I don't even see why it would be network connected. You could make it in such a way that the only method to view the video is to open the black box.

I have no doubt that it would cost millions. If it's important enough to do, it's important enough to spend extra on a (supposedly) reliable and secure system with some massive 24/7 maintenance contract. Unlike convenience store cameras that can just use off-the-shelf products and still deter criminals by just being there even when broken.

> Jeez, anyone could buy a security camera that recorded on a loop for a few hundred.

It probably isn't that easy, at least for the runways. Remember that Uber crash where they tried to convince everyone that the cyclist was near impossible to see? The dash cam used to record that video was standard, but also completely unable to handle even a well lit road at night. Good cameras and the weather proofing they need out in the open (they should still work after a snow storm) will cost a lot more, probably not infeasible but still a decent amount of cash.


> the weather proofing they need out in the open

He suggest to put one in the tower. There's no need to weather proof anything, or else poor controllers ;). In theory the tower see the runways well enough, or else the airport isn't well designed and the tower doesn't fill its own purpose.

> The dash cam used to record that video was standard, but also completely unable to handle even a well lit road at night.

Weren't we still able to see the cyclist? Weren't we able to boost the contrast to see her even better?

Sorry for the sarcasm, but even the worst camera show more than no camera at all... You can't argue that seeing a cyclist badly is worse than not seeing that same cyclist at all.

All it takes is a 100$ camera and you would get so much more evidence.


These cameras are everywhere. I've been doing this stuff in one way or another since 2008 or so. Did it for the Air Force (mounted on the tower) and for more commercial applications. It's a solved problem, and it is not particularly expensive.

Surely you've seen CCTV or other cameras in an outdoor enclosure before? You are being tracked from parking lot to boarding in some airports. We can put a few cameras up to watch the runways and taxiways.

https://airportimprovement.com/article/new-camera-system-ena...


It would be less useful than you’d think. Pilots are trained to articulate what they are doing and generally follow very well-practiced procedures. Another pilot can quite accurately reproduce what a pilot is doing just from the transcript, and the warning horns and other indicators are quite expressive of other things. The audio is good enough to hear breakers opening. Anything a video camera could see out the window that would be relevant will be remarked upon by pilots.

In the vast, vast majority of cases, yes. However, black boxes are only used in the minority where something has gone horribly wrong.

I've watched enough Aviation Disasters to know this isn't always good enough.

Which episode has a crash that (1) they failed to find a primary cause for and (2) would have been solved had they had video?

Dude, every single mishap since airplanes became reliable was pilot error at the root. It's never not pilot error. The FAA is a jobs program, similar to the old TVA, and note that all mechanics and pilots are still unionized. Governments have ensured that there will be no meaningful innovation in air breathing aviation, hence the space race...go ABOVE the bureaucrats.

Marc Andreessen describes his experiences seeking funding for Netscape as a disaster, particularly among the telecom giants of the age who had a potential stake in related businesses. Nobody thought the web or a web browser were worth investing in. Who would ever use those?

The moral of these stories is that originality scares the shit of people. Good ideas are not immediately apparent.


Reading Wikipedia, David Warren died in 2010, and his coffin was labelled "FLIGHT RECORDER INVENTOR DO NOT OPEN". Nice.

https://www.wsj.com/media/072310pod05.jpg


It kind of makes sense. If you had a high stakes job (human life, say a surgeon) and your boss wanted to record your every action, what would you think?

I think the FAA’s approach of “we do accident investigations so we learn, not to blame” went pretty far in making the idea more acceptable.


My idea was to make a lot of small transponders on the outside of a plane, that detach when in contact with seawater. And a few on long cables so that you could track that cable to the plane part. Those transponders could also have micro-sd cards with data on them.

fire is also another scenario.

magnetized steel wire as the recording medium.

I had no idea that was how recordings were (or at least could) be made at the time. Of course it makes sense-- early use of magnetic tape was something like nickel coated bronze, so there's no reason a wire couldn't also work.

Just a bit of electronics history I wasn't aware of that I thought interesting.



I think part of the reason for using wire was for fire proofing. The recording would be salvageable even after intense fires.

Later recorders used plastic tape but also have a water vessel. As the water boils it keeps the interior from going above 100C - until it boils away. The recorders are rated for maybe 30-60 minutes of direct exposure.

Recent recorders use solid state storage and, still, water to keep the internal temperature lower.


Tangental, but I wasn't aware of a ban on hobby radio activity during WW2 until I read this article. What would be the reasoning for that?

Probably a mix of: reserving supply of electronic components to military use; eliminating a possible cover story for spies; emission control to deny enemy navigation cues; limiting vectors for accidental information disclosure; creating an atmosphere of ‘We’re all engaged in this all out struggle together’; wartime bureaucratic expansion and overreach; and just generally more important things going on.

There’s an interesting aside in that neither a CVR nor a FDR would have been helpful in figuring out the cause of the Comet disasters. The way they did was brilliant.

Legal | privacy