Hacker News new | comments | show | ask | jobs | submit login
Ford disguised a man as a car seat to research self-driving (techcrunch.com)
339 points by gopalakrishnans 12 days ago | hide | past | web | 135 comments | favorite





It sounds goofy, but this strikes me as really clever. Compared to almost anything else with self driving cars adding a few indicators to outwardly show what it's going to do next (for pedestrians and other drivers) is a huge win.

There is a fair amount of waving for someone else to go first at a turn, seeing if a driver is distracted (on their phone) at a stop, making eye contact while using a cross walk etc that gets lost with SDVs.

Also, I really feel like they stole this idea from Raid 2 https://youtu.be/ErrRhXItBWc?t=118


* SPOILER *

That's how they made Knight Rider "driverless" scenes (http://www.knightriderarchives.com/knight_rider/season_one/g... <= can't find a clear shot, but if you find episodes online, it's pretty easy to see the double depth seat and raised up armrests)



How did I go through life without knowing that?

You are not alone & I am with you

I can't recall properly where I read it, but some time ago there was an idea to have stop signals also in front of the car, so the pedestrians would know if the driver/car is stopping.

There are also pedestrians who can take into account if the car is turning (obviously by watching the turning lights).

A self-driving car should have some reactions to people around it, so people could also respond properly to its actions, or know its intentions. There was a comment about an empty electric car near the intersection, staying still, and people around it can't tell if the car is parked, waiting, etc.


That seems a marvelously simple idea!

People understand what brake lights are and it's relatively simple to see the car slowing down that corresponds with the lights. Then assuming the driver doesn't use the hand brake then it should be obvious that the car is starting again when the brake lights go out.

All this flickering of lights seems way too complex


That won't work you can never know if the driver sees you or other cars, this is actually a very common problem at intersections. Drivers look for cars and slow down, people walk/bicycle over thinking the driver has seen them. This is how people die at intersections.

The slow "laptop in sleep mode" blink would be perfect for "I'm oblivious to the fact that I'm impeding traffic".

"waving for someone else to go first at a turn"

Ever since I heard of people getting blame for accidents, since they were effectively directing traffic (which only the police are allowed to do here), I've stopped waving people to go ahead and similar gestures, instead I'll give them space and wait for them to act of their own accord.

I've also stopped acting on other people's waves and gestures, I check for myself that everything is clear before I proceed.

Luckily I haven't witnessed an accident like this myself, but I know of people who've been T-boned on multi-lane streets because they were waved ahead by a guy in one lane, but someone came in another lane and couldn't see the turning car.


This annoys me too most of the time. Being in traffic is not the time to play Mr. Friendly, it's the time to follow the rules and thus behave predictably for the other drivers.

It's well known - a driver can only account for themselves, not you (or the environment).

As such, a driver that waves you on can only be trusted as saying "I will wait" but not "you are clear to go", you have to always verify this for yourself.


I don't mean to be offensive, but can someone please explain to me how the Stop signs work in US? Am I correct to understand that whoever arrived at the stop sign first goes first, regardless of where they are going? Is this considered a good system?

In EU you have strictly defined priority at stop signs - put simply, if you are going straight you have priority over someone turning, if you are turning, then the person turning right has priority over someone turning left. Simple as that.

I'm just thinking in context of automatic vehicles a strictly defined priority rules are easier to follow, no?


> I don't mean to be offensive

Reading this immediately makes me read the rest of your comment as much more negative and critical than I would otherwise. That's not necessary. If you want to ask a question about stop signs in the U.S., just ask it. (On the other hand, this is something you could google in about 5 seconds, so maybe your point really is just to criticize? Again, your opening line has thrown me off.)


Much like those notices that bear the heading "polite notice". I'll be the judge of that.

> In EU you have strictly defined priority at stop signs

Have you ever seen a 4-way stop in Europe?

I think what you try to explain is the rule for unmarked intersections.

> put simply, if you are going straight you have priority over someone turning, if you are turning, then the person turning right has priority over someone turning left. Simple as that.

That wouldn't work very well when people go straight in different directions...


(Disclosure: I learned how to drive in Germany, details may differ in other European countries, apart from the obvious mirroring when driving on the left)

The fallback rule for all situations is "yield to the right". If you're at an intersection with no signs or markings, you get to drive if there's no-one to the right of you.

This creates a deadlock when all roads are unmarked and there's a car on each of them at the exact same time but that almost never happens and can be resolved by slowly inching forward if the person to the right of you doesn't move and then exiting the crossing.

When entering from a road that has a solid line or a yield sign, you always have to yield to the traffic on the road you're merging into. A stop sign means you have to come to a full stop even if the road is empty and slowly approach until you can clearly see there's no traffic and only then merge (after yielding).

Most other rules follow from that.

I'm not sure what you mean by "4-way stop" but if there were a 4-way crossing with stop signs on all sides it would result in a deadlock and be resolved the same way. Except with the explicit restriction that everyone first has to come to a full stop and slowly move forward to determine the traffic situation.


In the US everyone stops and then whoever arrived first goes, breaking the deadlock in a well defined fashion (as long as everyone agrees on the order!).

>>Have you ever seen a 4-way stop in Europe?

I have actually.

>>That wouldn't work very well when people go straight in different directions...

Yes, I forgot to mention the "right hand principle" - in this case, whoever is on your right has priority over you. So if you are going straight, and the guy to your right is going straight, he has priority. The only situation where it doesn't work and leads to a deadlock is if you have a 4-way(or 3 way) intersection with equal-priority roads and everyone wants to go straight. Then it sort of reverts to this "whoever goes first wins" strategy - but that's incredibly rare and seeing roads with equal priority meet in this fashion is unusual.


> That wouldn't work very well when people go straight in different directions...

That was actually used in my parallel-programming course as the example of a deadlock caused by priority inversion. Of course it almost never happens that there are cars in all four directions at an intersection without traffic lights, so it isn't much of a problem in practice.


The parent's explanation didn't include the "yield to the right" rule, so it was problematic already with two cars.

It sounds goofy, but this strikes me as really clever.

Didn't the Top Gear / Grand Tour guys jokingly propose to do this?


Black Mirror had an episode where this went horribly wrong.

* SPOILER *

And Sherlock.


"Spoiler"? Oh, please, that one gets solved barely halfway through the episode.

Or this drive thru prank a year earlier:

https://youtu.be/xVrJ8DxECbg

I'm sure it had probably been done before that too. :)

Even better when you "put a robot in the driver's seat":

https://youtu.be/CdggQr05LAA


Thank you so much for this. Because of you I'm laughing uncontrollably in my cube at work now.

these are gold. thank you for the laughs.

People in the comments are making references to prank videos and film techniques, but this type of research study has also been conducted multiples times in the last few years by institutions such as Stanford [1], UC San Diego [2], Virginia Tech [3], and more.

[1] http://wendyju.com/publications/RO-MAN2016-Rothenbucher.pdf

[2] http://www.sandiegouniontribune.com/news/science/sd-me-ghost...

[3] https://www.citylab.com/transportation/2017/08/heres-the-rea...


I think the article is referring to the VirginiaTech study itself too

Curses! Shame on me for commenting before thoroughly reading the article.

Yep, these studies have been appearing at human-robot interaction conferences for a couple of years now.

A reporter apparently tried to talk to one of the drivers and they refused to acknowledge his presence: https://twitter.com/AdamTuss/status/894627339891609602

> and they refused to acknowledge his presence

Well of course, anything else would have been a car seat cosplay fail!


last thing that driver wanted to do was endanger their job. it was a very cush gig.

What an annoying reporter

He's actually a pretty decent reporter in the Washington DC area. He covers transportation, including all the failures of WMATA (DC metro).

"I'm with the news, dude."

This is hilarious to me for some reason.

"they just needed people to believe wholeheartedly they were using one."

Don't they mean "seeing one" instead of "using one"? Or am I missing something?

"fill in gaps where we currently communicate via subtle gestures, eye contact and other less obvious mechanisms."

This also struck me as odd - do pedestrians and other drivers really make "eye contact" with drivers to gauge their intention? I guess it's possible, but in many cases you can't see the driver's head, let alone eyes.


> do pedestrians [...] really make "eye contact" with drivers to gauge their intention?

Yes. Consciously, at every vehicle at an intersection, to make sure they aren't going to run me over. Same on a bike.

Other drivers? IDK, its harder through 2 windshields.


I do it all the time whether I'm walking, in a car, or on my motorcycle. If I look at them and see them looking at me it raises my confidence that they see me.

I've been given a ride in the passenger seat of a RHD car in a RHT country, ie the steering wheel is on the "wrong" side.

It's actually quite stressful before the driver (hopefully) stops at a zebra crossing after you accidentally caught eye contact with a waiting pedestrian staring at you.

The best strategy is to just look down, but that feels a bit stupid too when the pedestrians are clearly looking at you.


An electric car with an empty driver seat sitting at an intersection would be especially bad; when cross the street I would interpret that as a parked car, especially if it was by the curb (for either left or a right on a one-way street). Hopefully the car would be smart enough not to hit me but I don't want to have to start staring deeply at every parked car, trying to intuit if it is a robot about to pull away.

Do like in Canada: all cars must have their headlights on when driving (the lights automatically turn on).

edit: also, I feel like if we add new light signals for acceleration, for example, it would be nice to add that to all cars.


I've always wanted front brake lights for cars, that way as a pedestrian (or other driver) you can tell if a car is actively stopping as it's approaching you.

That's actually a really awesome idea. I wonder what the threshold of adoption would have to be before that signal is widely understood. Maybe that should become a necessary safety feature of driverless cars to replace the "feature" of being able to see what the driver is paying attention to through the window.

Also, it looks like there are multiple (??) patents on that idea: https://www.google.com/search?tbm=pts&hl=en&q=vehicle+front+...


As a cyclist, I sometimes see cars approaching crosswalks at a crawl, just creeping along at a few feet per second. I do that on my bike too, because I'm hoping the light will change before I have to come to a complete stop and put my foot down, but I have no idea why drivers do it, and it definitely doesn't feel safe crossing in front of a car like that.

In an automatic car, it's just about fuel efficiency and wear/tear. In a manual, it's that plus it's extremely annoying to stop and go at each semaphore.

Coming to a stop and just sitting there can warp the brake rotors. The area under the calipers doesn't cool off as fast as the rest of the brake rotor, and the localized heating causes the metal to bend.

You're supposed to creep forward a little, but some people might be overdoing it.


Not very safe, but a tiny bit more fuel efficient.

Brakes are a great way to turn momentum into heat. Then you have to spend fuel to regain momentum.

But, far less safe to move slowly when everyone else expected you were at a standstill.


Driverless cars won't do this... at least until we teach neutral networks frustration.

Those are daytime running lights. I wish headlights were mandatory, that would eliminate people driving in rainy twilight without lights on.

In California, you must have your headlights on if your wipers are on.

It does the opposite.

An always lit in vehicle display and DRLs that are just always on headlights makes people forget to turn their lights on when it's dark. This isn't a problem for them since they have their headlights on but drivers behind them don't see them from a distance at night because their tail lights are not lit up. This is particularly dangerous in traffic where everyone else has their lights on.

Making DRLs mandatory is tripping over a quarter to pick up a penny because it enables this behavior.


I just always turn my headlights on when I start my car. It doesn't matter if it's day or night, or it'll get dark while I'm driving, or if it starts raining or I go through a tunnel. The lights are just always on.

If everyone's car just did this automatically, the world would be a safer place. Why would you ever want the lights off?


Well, ask in countries which actually have that as law - Poland introduced it few years ago and the jury is still out on whether it made roads safer or not. I'd argue it didn't - mostly because as a driver your brain starts to associate lights = moving car. Which means that you start to ignore everything that doesn't have lights - bicycles, cars where someone forgot to turn the lights on. There were studies done on this and it's very inconclusive that it improves safety at all.

Given it's a safety feature the main reason not to do it would be if somehow it made things worse. Even if the effect is almost negligible you're still going to save a couple of lives per year.

Once LED lights are more standard and there's less issue of replacing bulbs or wasting electricity it seems a no brainer.


Well, that's the whole separate point in this discussion - normal headlamp bulbs are 60W - so you are using 120W of power to run two of them. How many cars are on the road for a country the size of Poland at a busy time of the day? 100k cars? That's 12 megawatts of power being used every day just to run some lights - and that's not counting rear lights at all. Generating 12 megawatts of power using thousands of ICEs is a massive waste of fuel, and it's probably incredibly hard to prove that conclusively, but I wouldn't be surprised if burning that fuel caused more deaths overall than having the lights on saves. I don't know if we can find any studies that would say either way though. Automotive LEDs are still within 10-15W range each, so while better than halogens they are not free.

And finally, I think there was a Polish study that said that yeah, while with the lights on you are more likely to notice a car, you are less likely to notice a cyclist - and hitting the latter has much worse consequences than hitting the former. So it might have made things worse.


Headlights and DRLs are not the same.

Most OEMs just make the headlights (and the headlights only, not the tail lights) always on in markets where DRLs are mandatory. The latter is what you get when you do what Canada did and require DRLs and don't require them to also be poor at being headlights.

Pretty much any time when you want to signal the intention of not going anywhere is a darn good time to have your headlights off. Cars actually have parking lights for exactly this reason. They also work great for creeping around your own driveway without throwing too much light on the windows of houses in the very early AM.


A lot of newer cars have an auto setting to handle this for you. DRL in daylight and headlights and tail lights as it gets dark or you go through a tunnel.

My last car rental, some moron previous renter turned the switch to DRL only instead of Auto. I did exactly as you described and drove with no tail lights and headlights that seemed dim but it wasn't my my car so it must just be that way. Eventually another driver flashed his brights from behind and I realized my mistake.


16 year olds would wire them to be on all the time.

When driving, if another driving appears to be doing something unusual I try to make eye contact to see what they're doing and try to determine their intentions.

I wonder if it's also beneficial to make eye contact since you have to turn your face towards the driver. Our brains are pretty good at picking out faces from a sea of visual information.

> do pedestrians and other drivers really make "eye contact" with drivers to gauge their intention?

I certainly do - most commonly when merging into a congested lane from a side street, stopping and making eye contact helps understand if everybody is paying attention (the merger can move forward and the merged is leaving a space open for them).

Motorcyclists also have a pretty large vocabulary (I'm not sure how many of these are actually used in practice): http://www.abate-il.org/Backroads/signals.html


Body language can convey a lot of information. We had to swerve onto the shoulder slightly yesterday to allow a motorcyclist room to move out of the way of a car in lane 2 who wanted to merge into lane 1, but didn't bother to signal or check his mirror. The wave we got from the cyclist afterwards said all of "thank you, and screw the guy next to me".

There's a rule of motorcycling that you look at a cars wheels, not the driver: "the car driver is likely to look you straight in the eye, before they pull out right in front of you".

Most of those are used when riding with a club or other group who know each other.

do pedestrians and other drivers really make "eye contact" with drivers to gauge their intention?

They do, and it's even advised by some. But I'm here to tell you that the eyes lie. I've had people look me directly in the eye and still pull out in front of me.

Look at the wheels. The wheels will always tell you what the vehicle is going to do.


When I'm jogging in the local park I'll often avoid making eye contact, as I think that this often works better to get people to stop (they aren't convinced you see them). Of course I'm prepared to stop if the vehicle doesn't.

I'm sort of intentionally being abrasive there though, as I think people drive too aggressively given the surroundings.


The real problem here is that cars are permitted in a park.

Maybe. It's a big park, 120 acres, and there is a hill on the non-lake perimeter that would make it difficult to construct adequate parking there.

>I'm sort of intentionally being abrasive there though, as I think people drive too aggressively given the surroundings.

This kind of passive aggressive behavior creates more problems than it solves.


I'm not dead yet.

Someone who's having a bad day is going to flip out on you for being an idiot and not paying attention (as they see it) when they have to stop short.

It'll probably happen eventually.

Looking like an oblivious idiot (which is indistinguishable from actually being one as far as other people care) who may dart out into the road at any second isn't going to make the world a better place. It will just increase the net frustration. Same goes for going the speed limit in the left lane. Roads are safer when people aren't angry.


"stop short"?

You are imagining much more aggressive behavior than I am actually engaging in. If it isn't clear that they have stopped, no way am I stepping in front of them.


> Look at the wheels. The wheels will always tell you what the vehicle is going to do.

This is exactly what I do, it's really obvious when the rim are rotating, even at very low speeds.

Can't say I look people in the eyes much while driving.


what about when they have spinners?

>"they just needed people to believe wholeheartedly they were using one."

>Don't they mean "seeing one" instead of "using one"? Or am I missing something?

the "they"s in the quote refers to the researchers, not "people"


> "they just needed people to believe wholeheartedly they were using one."

'they' is referring to Ford. Rephrased, "Ford just needed people to believe wholeheartedly [that] Ford [was] using one"


Pedestrians often do. For example, I'd want to make eye contact with a driver of a car attempting to make an unprotected left turn (and not looking in my direction) prior to crossing in front of it.

Drive.ai, (recently in the news for partnering with Lyft), has large LCD screens for signalling to pedestrians and other things. Probably a good idea.

This was actually in the news about a mysterious van with a driver pretending to be self driving.

https://www.theverge.com/2017/8/7/16109378/driverless-van-ar...


The local NBC channel (WRC-TV 4) covered it as well: http://www.nbcwashington.com/news/local/Driver-Dressed-Like-...

Watching the video it's almost impossible to tell what the flickering of the lights mean.

A much better idea seems the front brake lights mentioned by a couple of other commenters, plus a requirement for headlights if the car is on.


> This also struck me as odd - do pedestrians and other drivers really make "eye contact" with drivers to gauge their intention?

This statement both terrifies and piques my interest simultaneously ...


They(Ford) just needed people to believe wholeheartedly that they(Ford) were using one.

Ford was using the car. The people were seeing it. The second "they" in that sentence is referring to Ford, not the people.


I certainly try to make eye contact when walking in front of a car.

Yeah, it doesn't look like this disguise would work very well against someone sitting in the passenger seat.

For a second I thought they put a passenger in the "lap" of the driver - but looking at the pics there's no way...

do pedestrians and other drivers really make "eye contact" with drivers to gauge their intention?

The good ones do


And I suppose the "break lights" are lit when the CPU is nonfunctional?

It's interesting, everybody is thinking about how this was used for pranks and films, but I immediately thought of the much more sinister attempts to flee east Germany. I was unable to find a photo, but I remember seeing a car that was modified so a person could be sewn into the seat in a Berlin Museum.

Yes - it happened, also on the Mexico/US border, this has a guy as a car seat but leads with woman in a dashboard: http://www.snopes.com/photos/automobiles/dashboard.asp

I wonder how is it possible to find someone hidden in a dashboard? Were it dogs that found her?

I'm paraphrasing the news report because it's been a few years, but:

The border patrol agent became suspicious when the driver opened the glove box to retrieve his registration and a hand reached out from within the box to give it to him.


Sometimes, one can be too helpful.

This is really quite common. I read an article about a car in Pittsburgh where a driver was similarly disguised in order to research specifically other peoples reactions to driverless vehicles on the road. In the article I read, I believe it occurred in Pittsburgh, PA and the researcher was affiliated with a university (maybe Carnegie Melon?). The reporter was able to speak with the costumed researcher just briefly, but got a more substantive statement from the org doing the research.

That's literally the same project this article is about

Don Norman said in a talk they also have done that at his UCSD outfit.

Seems like they need to turn the cameras inward too (or has this been done already?); they can start observing "good" drivers providing signs to other drivers to then learn how to communicate a driving decision better.

It's kind of funny imagining a robot / self-driving car program going through driver's ed and taking the driver's license test with a human scoring it. Will car models and algorithms be evaluated like this too? :)


Really interesting study, would have liked to see some preliminary or even anecdotal findings discussed in the article.

I've seen some Youtubers pranking drive-through windows in the same way and they're pretty funny at the very least.


The setup reminds me of a pretty weird story called "The Human Chair": https://en.wikipedia.org/wiki/The_Human_Chair

I have been in a situation when I was mid-crosswalk when an idiot came screaming round a long shallow corner too fast (fast enough to have the white-line between his wheels).

I had a 50/50 choice. Keep going or run back. In the 3 or so seconds I had we made eye contact and I saw him return his eyes to my front (i.e. he was trying to get back on to his side of the road) so I moved backwards.

Not sure what I would have done if it were a self-driving car and unable to do the visual communication (Although the entire point might be moot since a self-driver wouldn't be in that situation)



So what were the results? Are pedestrians jerks? Do other cars cut them off?

New Turing test:

Is the driver in the car in front of you a human driver or a computer.


Looks like they put a cylon indicator on it? Not terrifying at all...

As for devising new signals which other drivers are supposed to watch to understand what the car plans to do, I'd say we need a lot more research to understand if this actually helps or just distracts.


Only successful if they can cross the Mexico border. :D

Prediction: this is the sell out costume for Halloween.

This article confused me a lot more before I realized Ford meant the company, not the individual.

Interestingly, it was the same for me. I was really confused for a few seconds until "self-driving" sank in and it became clear that Henry Fore could not have been alive for that.

In Henry's time, "self-driving" would probably be more likely used to describe a brick on the accelerator.

Also the first Ford that came to my mind was Ford from Westworld, not the carmaker.

MVP

So they re-did a prank from 2013 but for science [1]

[1] https://www.youtube.com/watch?v=xVrJ8DxECbg


The ol' car-seat-suit trick! That's how KITT from Knight Rider did it in 1982 as well! https://www.moviemistakes.com/entry91522

Disguising yourself as a car seat also features in this episode of 'Sherlock', https://en.wikipedia.org/wiki/The_Six_Thatchers

a programmer disguised as HM Aeron chair ... probably makes for more serene looking open floor office, and also minimizes the chances of the others bugging you with various stuff.

In London, with their right positioned drivers, more than once I got a scare of the phantom drivers. :-)

Not quite the promotion to deputy chair that he had expected.

Reminds me of the magician who goes through drive throughs in that costume.

I like the idea of an indicator showing what the car is 'doing'.


I can't wait until somebody patents "signaling pedestrians" and we all have to deal with brand-specific signaling mechanisms.

Thank Zod QWERTY was invented so long ago.

/cynicism


Off topic, but man do I hate looping gifs in my articles. It's so incredibly distracting.

Sure, play it once, maybe even twice. But an endless loop? That's just unnecessary.


It's not even very good communication. As there often is no clear start and end, when I scroll to the gif I have to watch it at least 2 or 3 times until I really know what is going on, where the loop starts and where it ends. Simple video would be so much better.

Leslie Nielsen did this forever ago: https://www.youtube.com/watch?v=xqJ3lx43lMg

The conspiracy theorist in me figures this kind of research will be used by companies that are falling behind in the self driving car race (no pun intended) to get governments to regulate against self driving cars.

Going off on a tangent here but is the HDR filter really necessary for those images? What is gained by using it? /rant

This isn't even the most elaborate market-research stunt I've seen.

Some of you older folk may remember an 80s sitcom called Small Wonder, which was a typical American suburbia family sitcom, except the youngest daughter, Vicki, was a robot. What I found unusual about the show was its copyright message: "©1987 The Small Wonder Joint Venture". In later years I found out that the show was produced in part for market research purposes, to solicit feedback about how Americans might feel about robots in the home -- especially robots that look and behave like humans. It was the 80s and we all thought that was coming Real Soon Now.


The unusual name ("The Small Wonder Joint Venture") is because the television show was the first product of a "joint venture" between 5 television stations [1], not because it was made by some secret home robotics manufacturer or marketing firm.

[1] https://en.wikipedia.org/wiki/New_Program_Group


Also, Booger produced 'wonder joints' to liven things up during the party scene in 1984's 'Revenge Of The Nerds'

https://youtu.be/WEp0fte3B8o


Interesting, I do recall that show, but can't find any corroborating info, sounds like a great urban legend though!

Are you denigrating my Teddy Ruxpin?



Applications are open for YC Winter 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: