Re:Google Driverless Car – The Obstacle Detection Unit
by vonmoltke on Monday September 08, 2014 @10:47AM
(#8285025)
Indeed. This is a sensor fusion problem I really want to work on. I loved the problem space while I was working for Raytheon, but I can't stand working on large government programs.
I'm surprised about the lack of IR sensors, though. I can think of cases where the IR data could be fused with data from other sensors, particularly visual, to improve object recognition. Particularly pedestrian and animal detection.
Re:Google Driverless Car – The Obstacle Detection Unit
by benjamincburns on Monday September 08, 2014 @03:25PM
(#8286659)
Would you care to share some of the new news? I'm presently looking for computationally efficient dense monocular VSLAM (visual simultaneous localisation and mapping) implementations to use in an upcoming personal project. openvis3d looks like it could be used to do binocular VSLAM, which would also be acceptable.
To seed the conversation, LSD-SLAM [1] looks really interesting, but last I checked the source has not yet been released. By "interesting" I mean that it's real time, dense, and runs on a low-powered (smart phone) CPU. I wish it was accurate enough to not need loop closure, but it's the best I was able to find over an afternoon of searching.
Re:Google Driverless Car – The Obstacle Detection Unit
by Gravityloss on Tuesday September 09, 2014 @07:22AM
(#8289578)
That looks very good. There are a lot of uses for SLAM and I hope the technology gets out of the labs soon. (My uses would be man portable indoor mapping of very complex structures with relatively good accuracy.)
Re:Google Driverless Car – The Obstacle Detection Unit
by ghaff on Monday September 08, 2014 @10:37AM
(#8284951)
Although that article also makes it unclear the degree of autonomy we're talking about here, using Lidar or otherwise. The big challenge, as noted in the link) is that there's this space between drivers still needing to drive (with assistance) and true driverless 99.999% just works. And within that gap, you potentially don't want to introduce certain automation because it will make drivers inattentive even though the machinery isn't quite up to the task of handling every contingency on its own. The billion dollar question is the timeframe for the 99.999% case--which enables a whole lot of radical use cases which the 99% case does not (such as no driver or no attentive driver). Just because the 99% use case could be here relatively soon doesn't mean that the 99.999% case couldn't be much further off (as in many decades).
Re:Google Driverless Car – The Obstacle Detection Unit
by bentcorner on Monday September 08, 2014 @12:38PM
(#8285716)
> which enables a whole lot of radical use cases which the 99% case does not (such as no driver or no attentive driver)
I'm excited for this. Car ownership and operation could be an interesting historical note 200 years from now. Family scheduling would become much more flexible and children incapable of operating a vehicle will be empowered to travel long distances.
Re:Google Driverless Car – The Obstacle Detection Unit
by ghaff on Monday September 08, 2014 @01:46PM
(#8286153)
I'm not convinced about car ownership. There are a lot of benefits in owning a car that you can keep "stuff" in. But just as Zipcars have presumably eliminated some car ownership use cases at the margins, fully autonomous cars would have a significant effect.
Re:Google Driverless Car – The Obstacle Detection Unit
by bentcorner on Monday September 08, 2014 @04:36PM
(#8287072)
> But just as Zipcars have presumably eliminated some car ownership use cases at the margins, fully autonomous cars would have a significant effect.
The big big thing I see about driverless cars is:
a) You can use it on-demand, even more easily than Zipcars. No need to physically be located near a lot, I can "hail" a driverless car and have it waiting for me when I need it. Smart routing and allocation can ensure that capacity is used smartly - it could drop somebody off while picking me up.
b) People who can't or won't drive can benefit from this. From young children to the elderly, transportation becomes much much easier. Have to take your kid to an appointment in the middle of the day? A driverless car can pick them up from school and bring them directly to the appointment and you can meet them there. You go to the appointment together, then driverless cars can take each of you back to your daily work, saving time all-around.
Yes - it's useful to have a car to keep stuff in. I can imagine you could keep one "on call" for a day out, and you can meet the car (at whatever exit is closest to you!) throughout the day to get supplies or drop things off.
Maybe I'm being too optimistic about what driverless cars can give us. Realistically I think hands-off driverless cars are within the next few years, but FUD and lawsuits will prevent more radical forward progress. It's certainly not unimaginable that it will take an entire generation of people to die off before truly autonomous cars are common and accepted.
Re:Google Driverless Car – The Obstacle Detection Unit
by sremani on Monday September 08, 2014 @03:51PM
(#8286821)
Car ownership on itself may have some benefits but most of the people who own car is not by choice. US cities are planned around cars. Walking and Biking are not only out of option for many, they are actually unsafe once you leave the neighborhood streets.
Re:Google Driverless Car – The Obstacle Detection Unit
by dragonwriter on Monday September 08, 2014 @04:42PM
(#8287115)
> There are a lot of benefits in owning a car that you can keep "stuff" in.
So, combine driverless cars with depots with loading/unloading robots, then you can leave stuff in your driverless car, and have access to it with the next driverless pickup.
Re:Google Driverless Car – The Obstacle Detection Unit
by bdamm on Monday September 08, 2014 @01:52PM
(#8286192)
Security is an interesting aspect as well. What's to stop someone from planting a location tracker on a shared car, then learning the locations of subsequent users? Or a bomb?
Re:Google Driverless Car – The Obstacle Detection Unit
by krapp on Monday September 08, 2014 @06:21PM
(#8287649)
The most likely novel vulnerabilities with these will be in the way they network with one another and municipalities. Also, doubtless people will try to carjack them or steal the hubcaps and stuff.
Re:Google Driverless Car – The Obstacle Detection Unit
by Thrymr on Monday September 08, 2014 @03:28PM
(#8286678)
Elon Musk thinks autopilot will come before driverless, too: http://www.bloomberg.com/news/2013-05-07/tesla-ceo-talking-w...
(or at least he did last year). I tend to agree. More cars have things like lane detection and adaptive cruise control that are steps towards autopilot. Truly driverless is not just a hard engineering problem, it is a huge unsolved social and legal problem.
Re:Google Driverless Car – The Obstacle Detection Unit
by ghaff on Monday September 08, 2014 @01:41PM
(#8286113)
Depends on the cases that aren't reliably handled by the computer. "Last 100 feet" problems associated with, say, parking could potentially be handled remotely. The bigger issue though is circumstances that require quick decision making by a competent human.
Re:Google Driverless Car – The Obstacle Detection Unit
by ddeck on Monday September 08, 2014 @12:32PM
(#8285675)
I'd be interested to hear from any experts in the are on this.
Why is Lidar so expensive? My vacuum cleaner has Lidar. Clearly not of the same grade as that on the Google car, but what is it that makes their unit cost tends of thousands of dollars and is this likely to fall dramatically at scale?
Re:Google Driverless Car – The Obstacle Detection Unit
by benjamincburns on Monday September 08, 2014 @03:01PM
(#8286545)
I don't have much direct experience with lidar, but with my background in in situ sensing I could probably land a job working with/on it. If I may be so bold as to presume that I just barely qualify as an "expert" here:
The lidar in your vacuum (presumably a neato-XV series; if so, you have fine taste sir/madam) is no more than a spinning laser range finder. These work by offsetting a laser from a linear image sensor (imagine a camera with a crazy wide aspect ratio, maybe 256x6 resolution) and calculating distance by simple trigonometry or a pre-calibrated look-up table. These are called spinning parallax lidar. Their output is two dimensional in polar coordinates (angle, distance).
More expensive lidar systems use carrier wave phase measurements (modulate the laser to a known signal, look for its reflection and calculate phase difference), direct time of flight measurements (like narrow beam sonic range finding, but with light, requires very expensive timing circuitry), doppler shift, and/or various structured light transformation/interference effects. Some units only use one of these techniques, some use a combination of these techniques. Most of these units operate using a number of lasers simultaneously to spit out 3 dimensional data, either as planar depth maps, or as 3D polar coordinates (angle-x, angle-y, radius). Either way, all of these are expensive because they require very fast and very precise complex analog circuitry, tight controls on noise sources (RF interference, temperature dependence, very good power isolation), tight controls on spurious emissions (GHz oscillators), careful calibration, and most importantly, very complicated software.
To me, time of flight is the most interesting, but the least practical/likely to be made inexpensive. For a DIY example of the kind of timing circuitry involved, check out [1].
That said, there's room to make it cheaper. One such way may be to use a structured light transformation technique similar to Kinect, but speed up the camera a bit and modulate the output to a known signal (gives better interference rejection).
Re:Google Driverless Car – The Obstacle Detection Unit
by randyrand on Monday September 08, 2014 @10:59AM
(#8285105)
Your questions both have the same answer (ie it's the same question) but that's a backwards way of thinking about it. You have to consider the cost of lidar at economics of scale, not before.
If your point is that lidar won't be cheap enough even at economies of scale to make a mass market device, then just answer OPs question directly. Your comment is just a restatement in a less intuitive way.
Re:Google Driverless Car – The Obstacle Detection Unit
by jbarrow on Monday September 08, 2014 @10:40AM
(#8284973)
I actually visited a German University (TU-Braunschweig) that works with a German car company to research driverless cars. The project that I saw [1] used Lidar as a verification method, but the ultimate aim was to be able to drive without it.
The interesting takeaway that I had from the visit was that the team working on this car believes that even partially autonomous cars are a technology that will come into the main stream only several decades into the future.
Re:Google Driverless Car – The Obstacle Detection Unit
by warfangle on Monday September 08, 2014 @10:47AM
(#8285028)
I wonder if they're using Lidar on their test vehicles in order to improve stereoscopic image analysis to the point where they can use cameras instead.
Re:Google Driverless Car – The Obstacle Detection Unit
by jre on Monday September 08, 2014 @11:29AM
(#8285299)
As far as I know, stereo vision has some problems with uniform surfaces (snow-covered road). Since any two points on a uniform surface will have similar descriptors, it's pretty hard to match them.
There are some solutions to this problem (using a Markov Random field or some assumptions about the surface), but I'm not sure how reliable they are.
Re:Google Driverless Car – The Obstacle Detection Unit
by trhway on Monday September 08, 2014 @03:09PM
(#8286576)
we have pretty high retina resolution. To feel the limits of our stereo machinery, look at any monotone surface with monotone lightning at the distance that you can' distinguish any local mini-features of the surface - looking at the smooth ceiling of something like this - and you can feel how your eyes strain trying to find (and can't) the correct focus distance to distinguish the features and thus to get stereo working by matching the features from the left and right channels. The same feeling can be felt when you look at say vertical bar patterned surface so that eyes/brain have harder time to match specific bars from left and right eye. In both cases the surface should be large enough to cover the focus spot in your vision space, so that eyes/brain couldn't get help from relative position of the surface vs. other objects.
Those are actually the same issues one encounters when develops computer stereo :) The modern cameras got to the retina resolution and computer power today is able to do stereo match on those resolutions - this is why we started to get meaningful results. The computer stereo will surpass human's because cameras can have higher resolution, higher sensitivity and they can see in different wavelengths in addition to visual. Plus you can "paint" the objects ahead of you with IR for example, so the stereo match would be even easier.
Re:Google Driverless Car – The Obstacle Detection Unit
by jre on Tuesday September 09, 2014 @02:36PM
(#8291831)
In addition to trhway's answer, I'd say we use a lot of assumptions about the world : houses usually have flat, orthogonal walls, streets are flat, etc.. Of course, we can have the computer learn those assumption, but it's more complicated than pure dense stereo.
Re:Google Driverless Car – The Obstacle Detection Unit
by Symmetry on Monday September 08, 2014 @10:23AM
(#8284880)
I'm curious how you would detect debris on the road at speed reliably without it. I wouldn't trust a visual system to distinguish obstacles from discolorations, and other systems wouldn't have the precision or range.
Re:Google Driverless Car – The Obstacle Detection Unit
by vonmoltke on Monday September 08, 2014 @10:32AM
(#8284931)
It can be done with an RF system using SAR techniques, if the vehicle has a precision INS installed. Fuse that with data from other sensors, and you can build a pretty good topo map of the road ahead of the vehicle.
Re:Google Driverless Car – The Obstacle Detection Unit
by zo1 on Monday September 08, 2014 @12:11PM
(#8285566)
I genuinely don't know all three acronyms you used there.
Be honest. Do you think that most of the people that are browsing over and reading these comments are knowledgeable enough in your niche field to know all those acronyms?
Re:Google Driverless Car – The Obstacle Detection Unit
by vonmoltke on Monday September 08, 2014 @03:31PM
(#8286701)
SAR -> Synthetic Aperture Radar
INS -> Inertial Navigation System
RF -> Radio Frequency
Yes, I should have spelled out the first two. The third I do expect most people here to know, given the context of the thread.
> Do you think that most of the people that are browsing over and reading these comments are knowledgeable enough in your niche field to know all those acronyms?
RF and sensors may be a portion of electrical engineering that is rarely discussed here, but it is hardly "niche".
I made a quick post where I should have made a more detailed one. Mea culpa.
Re:Google Driverless Car – The Obstacle Detection Unit
by zo1 on Monday September 08, 2014 @01:20PM
(#8285961)
"RF" was the third one. And I suppose I could take a guess and say it means "Radio Frequency". But just because the acronym's letters fit the phrase, doesn't mean I'm right as I am not knowledgeable in the field being discussed.
Re:Google Driverless Car – The Obstacle Detection Unit
by Symmetry on Monday September 08, 2014 @09:31AM
(#8284580)
I wonder if the radars they're using can do frequency binning, that is figuring out the relative velocity of objects they detect via Doppler as well as their distance?
Also, that 64 beam 360 degree lidar makes me drool a little bit.
Re:Google Driverless Car – The Obstacle Detection Unit
by vonmoltke on Monday September 08, 2014 @10:00AM
(#8284722)
I don't see why not. Pulsed transmitters are fairly ubiquitous nowadays. A bit more complicated and expensive than CW transmitters, but easy nonetheless.
Re:Google Driverless Car – The Obstacle Detection Unit
by sharemywin on Monday September 08, 2014 @09:40AM
(#8284626)
This is one of the places where I think there should be colaboration across companies. I'm not sure "not hitting pedestians" should be market differentiator.
Re:Google Driverless Car – The Obstacle Detection Unit
by mrec on Monday September 08, 2014 @11:41AM
(#8285372)
It's not a binary checkbox feature, though. I doubt any implementation will make it utterly impossible to hit a pedestrian - and that's fine, as long as they do better than meatbag drivers. A better detection/avoidance implementation is absolutely a differentiator, mostly via lower insurance premiums for operators.
Re:Google Driverless Car – The Obstacle Detection Unit
by JoeAltmaier on Monday September 08, 2014 @11:48AM
(#8285408)
That's not clear, in our litigenous society. I hit a pedestrian: I get in trouble. A driverless car hits a pedestrian: all driverless cars (of that manufacturer) are potentially 'at fault' and recalled. Its definitely a different bar to pass.
Re:Google Driverless Car – The Obstacle Detection Unit
by josho on Monday September 08, 2014 @06:03PM
(#8287580)
That's interesting. A computer hits a pedestrian, then presumably the algorithms are updated and the probability of that scenario playing out again drops to 0. While A human hits a pedestrian, and that scenario is likely to never occur for that driver again, but for all other drivers the probability remains the same.
Re:Google Driverless Car – The Obstacle Detection Unit
by krapp on Monday September 08, 2014 @06:08PM
(#8287595)
Although you technically may be right (depending on how easy 'updating the algorithm' turns out to be - it being as easy as updating an app seems a bit on the crazy side to me) I don't know that the likelihood for all human drivers of hitting a pedestrian is necessarily high, though.
It exists, obviously, but I think human drivers already come with the "don't hit pedestrians" firmware installed, and it mostly seems to work.
To me it won't be demonstrable that driverless cars are vastly superior until there are a decent number of them on the streets, mixing it up with traffic in all kinds of weather, dealing with unexpected hazards and inaccurate maps, for an extended period of time.
Re:Google Driverless Car – The Obstacle Detection Unit
by mrec on Monday September 08, 2014 @02:49PM
(#8286475)
But recalls happen already, for all sorts of reasons. They're expensive, and they can have a big impact on a manufacturer's reputation, but they don't mean that nobody can sell a car in the first place.
Re:Google Driverless Car – The Obstacle Detection Unit
by robbrown451 on Monday September 08, 2014 @11:44AM
(#8285384)
It's not a black and white thing. No driverless car, nor human driver, can claim to 100% avoid accidents, even those that are theoretically avoidable. But some can be better/safer than others.
Most likely, though, the way competition will play out around this issue is how conservative the decisions the vehicle makes will be. So a "lesser" vehicle might come to a complete stop at a crosswalk, while a better one (as well as any reasonable human driver) might be able to calculate that it can safely drive on through because that pedestrian is well way out of the way. Both vehicles can be equally safe, but one gets you to your destination a lot faster because it is better at accurately estimating collision risk, so it doesn't have to compensate by sacrificing timeliness.
Re:Google Driverless Car – The Obstacle Detection Unit
by josho on Monday September 08, 2014 @06:16PM
(#8287629)
Herein lies why I can't wait for driverless cars. Cars are supposed to stop while a pedestrian is in an intersection, no matter if the pedestrian has already passed the car and is a 'safe distance away'. The reason is by stopping you are signalling to other drivers that there may be a pedestrian in the intersection. Just in case the other drivers don't see the pedestrian for some reason.
I agree with your point however, that competition is important on performance elements like the quality of the drive.
Re:Google Driverless Car – The Obstacle Detection Unit
by randyrand on Monday September 08, 2014 @11:24AM
(#8285266)
It's a gut instinct to agree, but I think you underestimate the power of competition.
There invariably should be partnerships. But if there was only a single driverless car solution, what prevents companies from getting complacent? Why aim for more than "good enough?" Why put 100 billion into a project if it won't help you sell more cars than your competitors?
Its seems odd that anyone would go with a provably less safe driver less car solution, but perhaps some systems are better at other types of driving than others? At the very least, there has to be legislation that defines a bare minimum (some multiple of human competence or maybe the 1 percentile), but beyond that, competition is more or less necessary.
Re:Google Driverless Car – The Obstacle Detection Unit
by TeMPOraL on Monday September 08, 2014 @09:43AM
(#8284640)
Probably a bit, yes, but there are ways to filter out the interference to some extent. Anyway, I'm strongly hoping that they'll make cars that talk to each other - so instead many cars adding noise to each other's readings, we'll have cars that share data and dynamically delegate the task of detailed region mapping to few cars that would not interfere with each other.
Re:Google Driverless Car – The Obstacle Detection Unit
by TeMPOraL on Monday September 08, 2014 @11:28AM
(#8285291)
Probably in the same way you handle malicious actors now - people deliberately obstructing view of other cars, shining laser pointers at drivers, etc. A very important observation is that those are extremely rare events.
In a way, humans are also machines, and we have very similar attack vectors, so just extending our current solutions might be good enough to accommodate self-driving cars on our roads.
Re:Google Driverless Car – The Obstacle Detection Unit
by aetherson on Monday September 08, 2014 @12:58PM
(#8285821)
I think that the way you handle malicious actors now is that people know that they have a laser in their eyes. This would be more like "what if someone could just make another car or obstacle invisible to you, and you have no way of knowing it."
I agree it is a very rare event, and will always continue to be. But it would need to be handled, and handled in ways other than, "Hi, here's a 100 car pile-up."
Similarly, and probably less rarely, you'd need to handle scenarios in which another actor wasn't malicious per se, but faulty to a degree that it was functionally malicious. Of course, we've always had to handle the idea that somebody might be driving a car with bad brakes, and that's dangerous enough, but before now a car with bad brakes wouldn't also make the brakes bad for ten cars around them. I think that's worthy of some contemplation.
Re:Google Driverless Car – The Obstacle Detection Unit
by TeMPOraL on Monday September 08, 2014 @04:12PM
(#8286951)
I agree with the sentiment. There's a lot to consider and test in practice, though any of those scenarios don't sound like show-stoppers. After all, humans are also machines, and where a person can "know that they have a laser in their eyes", a machine can equally well realize that one of its sensors suddenly started reporting weird results that don't match the expectations and that don't seem to agree with other sensors onboard or with what other cars nearby are reporting.
We'll have to sort out a lot of those scenarios, many by means of "testing in production", but as long as self-driving cars can outperform humans on average, we'll be better off.
Re:Google Driverless Car – The Obstacle Detection Unit
by meowface on Monday September 08, 2014 @10:00AM
(#8284720)
That's a good question, and it makes me think that for that reason, Google probably won't actually be able to implement what the above post is suggesting.
The most that can happen is each car will probably end up sending Google certain diagnostic information, but actual environmental sensing will have to be done by the car itself, with maybe some help from aerial cameras or satellites that Google controls.
Re:Google Driverless Car – The Obstacle Detection Unit
by netrus on Monday September 08, 2014 @11:01AM
(#8285118)
I don't know, is this really an important scenario? If I wanted to harm a specific individual, I would manipulate the individuals car (like I would today). And who wants to drive an undirected attack on everyone close? That would be considered a quite serious crime that takes a substantial technical effort, with little to gain.
Re:Google Driverless Car – The Obstacle Detection Unit
by jonlucc on Monday September 08, 2014 @10:34AM
(#8284942)
Maybe by voting? It probably shouldn't be used for the most vital aspects, but it would be less likely to have 5 or 6 malicious cars around than just 1.
Re:Google Driverless Car – The Obstacle Detection Unit
by robbrown451 on Monday September 08, 2014 @11:51AM
(#8285439)
I understand how the cars can be aware of other objects on the road, but are they actually aware of themselves and where they fit into the environment? It seems they'd have to be in order to function, right?
So we've get self-aware machines now. Did anyone even notice?
Re:Google Driverless Car – The Obstacle Detection Unit
by Rangi42 on Monday September 08, 2014 @12:42PM
(#8285741)
If it were programmed to keep itself clean, and had an algorithm to do that ("scrape the dirty area against a shrub"), I see no reason why it couldn't pass the mirror test.
Re:Google Driverless Car – The Obstacle Detection Unit
by Rthakur on Thursday September 11, 2014 @11:59AM
(#8303326)
Velodyne recently released a 16 channel LIDAR for ~ $8000. I agree with you that the cost can come down tremendously with volume. You also need to consider the differences within LIDAR systems. The velodyne system is for 360 degree field of view. Other automotive LIDAR systems (See continental, Valeo, Ibeo, etc) with lower FOV and functionality (forward collision warning, blindspot detection,etc) are much lower cost > 10x ....
So actually this might be a good fit for Google after all. Sensors are getting trivialized.
I'm sure Google is working for a LIDARless car. They've published their stereo vision to 3d library.
https://code.google.com/p/openvis3d/
I'm surprised about the lack of IR sensors, though. I can think of cases where the IR data could be fused with data from other sensors, particularly visual, to improve object recognition. Particularly pedestrian and animal detection.
Have you thought about working for Tesla? [email protected]
To seed the conversation, LSD-SLAM [1] looks really interesting, but last I checked the source has not yet been released. By "interesting" I mean that it's real time, dense, and runs on a low-powered (smart phone) CPU. I wish it was accurate enough to not need loop closure, but it's the best I was able to find over an afternoon of searching.
1: http://vision.in.tum.de/research/lsdslam
German car makes (Mercedes, Audi, BMW,...) think they can create a driverless car without the need for expensive Lidar, see for example http://spectrum.ieee.org/cars-that-think/transportation/self...
I'm excited for this. Car ownership and operation could be an interesting historical note 200 years from now. Family scheduling would become much more flexible and children incapable of operating a vehicle will be empowered to travel long distances.
The big big thing I see about driverless cars is:
a) You can use it on-demand, even more easily than Zipcars. No need to physically be located near a lot, I can "hail" a driverless car and have it waiting for me when I need it. Smart routing and allocation can ensure that capacity is used smartly - it could drop somebody off while picking me up.
b) People who can't or won't drive can benefit from this. From young children to the elderly, transportation becomes much much easier. Have to take your kid to an appointment in the middle of the day? A driverless car can pick them up from school and bring them directly to the appointment and you can meet them there. You go to the appointment together, then driverless cars can take each of you back to your daily work, saving time all-around.
Yes - it's useful to have a car to keep stuff in. I can imagine you could keep one "on call" for a day out, and you can meet the car (at whatever exit is closest to you!) throughout the day to get supplies or drop things off.
Maybe I'm being too optimistic about what driverless cars can give us. Realistically I think hands-off driverless cars are within the next few years, but FUD and lawsuits will prevent more radical forward progress. It's certainly not unimaginable that it will take an entire generation of people to die off before truly autonomous cars are common and accepted.
So, combine driverless cars with depots with loading/unloading robots, then you can leave stuff in your driverless car, and have access to it with the next driverless pickup.
Why is Lidar so expensive? My vacuum cleaner has Lidar. Clearly not of the same grade as that on the Google car, but what is it that makes their unit cost tends of thousands of dollars and is this likely to fall dramatically at scale?
The lidar in your vacuum (presumably a neato-XV series; if so, you have fine taste sir/madam) is no more than a spinning laser range finder. These work by offsetting a laser from a linear image sensor (imagine a camera with a crazy wide aspect ratio, maybe 256x6 resolution) and calculating distance by simple trigonometry or a pre-calibrated look-up table. These are called spinning parallax lidar. Their output is two dimensional in polar coordinates (angle, distance).
More expensive lidar systems use carrier wave phase measurements (modulate the laser to a known signal, look for its reflection and calculate phase difference), direct time of flight measurements (like narrow beam sonic range finding, but with light, requires very expensive timing circuitry), doppler shift, and/or various structured light transformation/interference effects. Some units only use one of these techniques, some use a combination of these techniques. Most of these units operate using a number of lasers simultaneously to spit out 3 dimensional data, either as planar depth maps, or as 3D polar coordinates (angle-x, angle-y, radius). Either way, all of these are expensive because they require very fast and very precise complex analog circuitry, tight controls on noise sources (RF interference, temperature dependence, very good power isolation), tight controls on spurious emissions (GHz oscillators), careful calibration, and most importantly, very complicated software.
To me, time of flight is the most interesting, but the least practical/likely to be made inexpensive. For a DIY example of the kind of timing circuitry involved, check out [1].
That said, there's room to make it cheaper. One such way may be to use a structured light transformation technique similar to Kinect, but speed up the camera a bit and modulate the output to a known signal (gives better interference rejection).
1: http://piotr.nikiel.info/?page_id=95
If your point is that lidar won't be cheap enough even at economies of scale to make a mass market device, then just answer OPs question directly. Your comment is just a restatement in a less intuitive way.
The interesting takeaway that I had from the visit was that the team working on this car believes that even partially autonomous cars are a technology that will come into the main stream only several decades into the future.
[1] http://www.spiegel.de/video/fahrerloses-auto-leonie-kreuzt-d...
Sorry, the link is to a video in German. I'm sure that if you search for Braunschweig "Leonie" you may find a source in English.
There are some solutions to this problem (using a Markov Random field or some assumptions about the surface), but I'm not sure how reliable they are.
ed. I guess that's the trillion dollar question...
Those are actually the same issues one encounters when develops computer stereo :) The modern cameras got to the retina resolution and computer power today is able to do stereo match on those resolutions - this is why we started to get meaningful results. The computer stereo will surpass human's because cameras can have higher resolution, higher sensitivity and they can see in different wavelengths in addition to visual. Plus you can "paint" the objects ahead of you with IR for example, so the stereo match would be even easier.
Be honest. Do you think that most of the people that are browsing over and reading these comments are knowledgeable enough in your niche field to know all those acronyms?
Probably not.
Yes, I should have spelled out the first two. The third I do expect most people here to know, given the context of the thread.
> Do you think that most of the people that are browsing over and reading these comments are knowledgeable enough in your niche field to know all those acronyms?
RF and sensors may be a portion of electrical engineering that is rarely discussed here, but it is hardly "niche".
I made a quick post where I should have made a more detailed one. Mea culpa.
RF is probably radio frequency
INS is probably inertial navigation system - http://en.wikipedia.org/wiki/Inertial_navigation_system
> driverless car (i.e. not "self-driving car")
> robotic chauffeur
Is it me, or the article is pushing some new terminology?
Also, that 64 beam 360 degree lidar makes me drool a little bit.
It exists, obviously, but I think human drivers already come with the "don't hit pedestrians" firmware installed, and it mostly seems to work.
To me it won't be demonstrable that driverless cars are vastly superior until there are a decent number of them on the streets, mixing it up with traffic in all kinds of weather, dealing with unexpected hazards and inaccurate maps, for an extended period of time.
Tesla is already updating the firmware of their cars over-the-air[1]. It seems likely that driverless cars will have the same feature.
[1] http://www.theverge.com/2012/9/24/3385506/tesla-model-s-over...
(but still crazy.)
Most likely, though, the way competition will play out around this issue is how conservative the decisions the vehicle makes will be. So a "lesser" vehicle might come to a complete stop at a crosswalk, while a better one (as well as any reasonable human driver) might be able to calculate that it can safely drive on through because that pedestrian is well way out of the way. Both vehicles can be equally safe, but one gets you to your destination a lot faster because it is better at accurately estimating collision risk, so it doesn't have to compensate by sacrificing timeliness.
Competition is important.
I agree with your point however, that competition is important on performance elements like the quality of the drive.
There invariably should be partnerships. But if there was only a single driverless car solution, what prevents companies from getting complacent? Why aim for more than "good enough?" Why put 100 billion into a project if it won't help you sell more cars than your competitors?
Its seems odd that anyone would go with a provably less safe driver less car solution, but perhaps some systems are better at other types of driving than others? At the very least, there has to be legislation that defines a bare minimum (some multiple of human competence or maybe the 1 percentile), but beyond that, competition is more or less necessary.
In a way, humans are also machines, and we have very similar attack vectors, so just extending our current solutions might be good enough to accommodate self-driving cars on our roads.
I agree it is a very rare event, and will always continue to be. But it would need to be handled, and handled in ways other than, "Hi, here's a 100 car pile-up."
Similarly, and probably less rarely, you'd need to handle scenarios in which another actor wasn't malicious per se, but faulty to a degree that it was functionally malicious. Of course, we've always had to handle the idea that somebody might be driving a car with bad brakes, and that's dangerous enough, but before now a car with bad brakes wouldn't also make the brakes bad for ten cars around them. I think that's worthy of some contemplation.
We'll have to sort out a lot of those scenarios, many by means of "testing in production", but as long as self-driving cars can outperform humans on average, we'll be better off.
The most that can happen is each car will probably end up sending Google certain diagnostic information, but actual environmental sensing will have to be done by the car itself, with maybe some help from aerial cameras or satellites that Google controls.
Google is utilizing multiple techs in this invention.
So we've get self-aware machines now. Did anyone even notice?
[1]: http://www.instructables.com/id/Line-following-Robot-with-Ar...
If someone wanted to sit around and just produce lidar units, i'm sure it could be done for 100x less.