Slashdot Mirror
Autonomous Race Cars
Octothorp writes: "Though not as complicated as the underwater
vehicles. There is an annual competition sponsored by National Semiconductors to build an autonomous race car. They move along pretty well too, at almost 9 ft/s. More technical information on how they are built is available on a Berkeley page, and there's a video of the winning run for 2002."
I recall an experiment done in California a few months back whereby cars automatically drive themselves on a stretch of a highway equipped with magnetic strips down the centre of the lane.
Since public transport seems to be out of the question for medium-distance transportation in the States (witness Amtrak's plight - and CoachUSA's financial gymnastics of late) automated private transport might fill the gap - provided price can be brought down.
It is really interesting to see what comes out of this. Certainly more applicable in the near future than football-playing robots - or this terrible series in the UK with robots battling each other, armed with saws, hammers and whatnot contraptions..
Curious,
Michel
Fedora Project Contribut
Just to clarify for those who made the same initial assumption I did.
------------------------
Jack not name, jack job!
Linking to a video file from the frontpage of /.?
Wowzers! Methinks flaming shards of this server will be moving - autonomously - at 9 ft/sec very shortly...
[/bad joke]
The key to this particular competition seems to be the size and weight limits on the cars. This is a contest for little RC-style cars. The technology exists to go much faster, but not in this form factor.
Surprise surprise, the site is dead now...
It would be interesting to take one those cars (with the electronics still on for weight fairness etc), install some RC gear, and see who's faster.
I suspect that the human would still be the better driver/faster at the moment. Also, a human driver can navigate traffic. These guys look like they might have a hard time in a race with other cars ;)
this might give me the skills to leave the B-mains and hit the A-mains. Assuming they let me tape their ozite carpet. :p
There are 5280 ft in a mile, 3600 seconds in an hour *9 = 32400 FPH /5280 = 6.136...MPH
Why even post iit in FPS when you know almost every person that reads it will have to do the math, why don't you just give it in nanometers per light year and let us go from there?
Saying Java is nice because it works on all OS's is like saying that anal sex is nice because it works on all genders.
It's my opinion that autopilot for your car is an inevitable development. It will become the only way to keep highways scaling. Adding more lanes to add capacity works to a point, but doesn't work forever. I don't think it's feasable to expect a human to navigate a twenty or thirty lane highway. And the density of traffic which an existing highway can carry is limited by the poor driving ability of humans.
I think eventually high-capacity highways will require the use of an autopilot. Doing so would allow the cars to be run with inter-car gaps which would be suicidal with a human behind the wheel. Most stop-n-go situations are due to bad planning on the part of drivers. They speed up too much when traffic clears ahead, zoom up on the cars in front, and then have to slow down to avoid an accident. This type of driving creates waves of congestion which travel backwards down the highway, and is due entirely to poor coordination among drivers. But there is no reason that under computer control rush hour can't cruise along at 60 miles an hour with a car length or less between cars. I bet you could easily triple the maximum capacity of a highway, not to mention getting everybody to their destination faster and with better fuel economy to boot. The R&D will be expensive, but like any electrics the hardware will be practically free once developed. Compared to the cost of expanding existing freeway's, it will make sense finacially too. I figure it's maybe 10 years out.
American's are not likely to give up their cars for any sort of public transportation, no matter how impractical cars become with rising fuel costs, increasing travel times due to congestion, increasing insurance rates, etc. Most Americans have convinced themselves that they enjoy sitting in stop and go traffic, as long as it's in a car and not a bus. But if we could figure out a way to let them keep their cars, reduce pollution, reduce accidents, let them safely talk on their cell phones, and not have to build mile-wide highways I suspect a lot of us might go for it.
The streets are jam-packed with thousands of vehicles travelling in irregular serpentine patterns.
None of these machines (except mine, of course) contain any type of human intelligence, but it's interesting to watch the AI at work. At night the roads look like Conway's Game of Life running on a computer with bad RAM.
Expert systems allow some vehicles to negotiate left turns from right lanes and to outbrake school busses when entering a rotary.
Fuzzy logic is essential for speed control, stop lights, parking and many other mission-critical tasks.
Genetic algorithms tend to select the maneuvers which are least expected by other vehicles.
Task scheduling is done according to driver convenience. For example, turn signals are always lower priority than dialing a call on the mobile phone.
Most communications between vehicles is a crude form of "digital" communication.
Unfortunately, most of the vehicles are Windows-based which results in a high rate of crashes. Mack trucks seem to be better than average.
With a good onboard computer, the car could build a model of the track as it went around, and calculate the optimal path & speeds to use on all subsequent laps. Using lasers or ultra sonic distance sensors would let the robot know when a turn was going to happen a lot sooner the the few inches of warning it gets in their setup. If you put an accelerometer in the car, you could even have it self-calibrate, discovering it's own acceleration curve, maximum lateral acceleration, and braking. It could then use those values to find the perfect path through the track on the second lap.
Yes, it sounds like a good idea, and in fact that was the original plan for the winning car. But it sounds a lot easier than it is. First of all, time constraint wise - these cars are built from scratch, and tuned to perform well in under fifteen weeks. (Yes, that includes all the sensor circuits, and power supply electronics) Secondly, there is a major problem with wheel slippage - if your wheels slip, you don't know where you are anymore.
A entry from 2001 went slow around the track the first time to memorize it, and then used that information on the second round in order to predict turns and change speed. It used the track crossing location to resync where it thought it was on the track. But if you look at the track layout, there are large section with no track crossing. i.e. wheel slippage - knowing where you actually are - is the main problem to be solved for memorization type approaches.
You can't see it in the video, but the winning 2002 car does detect and speed up (slightly) on straight aways.
Steve VanDeBogart
I hope I miscalculated this, but I figured 9ft/s is around 6mph. 9ft/1s:1mi/5280ft:3600s/1hr For those of you in modern nations, I apologize for the Imperial units. However; I am a speed-crazed American and albeit autonomous, if it can't go 45mph, I'm not impressed...
On a more serious note: Is the direction finding logic so intense that a greater speed can't be safely achieved?
Maybe we'll have auto pilots in 100 years, but I don't see it happening in my lifetime. Yeah, an good autopilot could let cars run closer together, under normal circumstances, but if anybody gets a flat, people are gonna die. A computer just can't spot all the road hazards a person can. It's not going to know about that cinderblock that fell off a truck, or the deer that's running into the road.
Americans don't use public transportation, because public transportation here sucks.
Ride a bus in Germany, then ride a bus in NY. There's a world of difference. For me to use public transportation more often, it needs to become less dirty and less dangerous. For me personally, it would also need to be going 24/7.
One might also care to realize that things are laid out differently in America than they are in Europe. Ex: If I want to buy some milk, it would take me 40 minutes to walk to the closest place, and another 40 minutes to walk back. There is no mass transit I could take there even if I wanted to. America is BIG. I can get on the highway, drive for 3 hours and still be in the same state.
I think cars are impractical in certain areas (NYC, Boston, etc). In most places, they are a necessity. Where I live traffic is very efficient. I'm lucky enough to live in an area when most people actually know how to drive, and near a highway big enough to handle their traffic. I love driving on it. Even during rush hour, the average speed is 70.
Life is too short to proofread.
Uh. This is a comment coming from someone who obviously hasn't done much programming. You want to design the system that accounts for real life racing circumstances? Coincidently, I'm watching Driven right now, a very bad movie. But it happens to drill the point into my mind that racing is nothing short of chaos. You argue that game AI is sufficient to use in real life. You're basically saying that the AI can deal in a chaos environment. However, the computer generates the chaos. It knows what each car is doing and is planning on doing. Each car object in the game is interwoven with the other objects. Real life is not this way. You can't predict what the other cars are doing. Try designing code to react to such a complex environment in which people can easily die if things don't work *perfectly*.
"Nature doesn't care how smart you are. You can still be wrong." - Richard Feynman
This is a very simple control problem to solve... I actually think that you could do best in this competition by working on making the car faster; Most of the cars are going to have similar line following abilities.
Actually, that's not the case at all. If I remember correctly, the motor is capable of doing something in the 30 MPH range (though probably not for too long). It turns out that working on the control problem, and the accuracy/reliability of inforation you're working with is much more important.
Steve VanDeBogart
My son is building one of these. He gets the parts from a magazine subscription. Light following/repulsion, sonar detection, line following. And seems to go pretty fast to boot.
MrCreosote Meow!Thump!Meow!Thump!Meow!Thump! "You're right! There isn't enough room to swing a cat in here!"
The winner was a constant-speed car. At 6MPH. Cars aren't learning the track and then driving it at a good speed; they're just dumb line-followers.
Battlebots are much cooler.
... why, that's almost sixteen and a half kilofurlongs per fortnight!
yes, we have no bananas
You just can't kludge up something like that in a few weeks. Anybody who's done power electronics looks at those numbers in awe. Plus they come with features like total self-protection against dead short, overheating, reverse polarity, etc. And all for about $100. Industrial motor controllers with comparable ratings cost thousands and are huge.
I have no real problem with "feet per second" as a measurement. But, out of the hundreds of articles that Timothy gets to choose from every day, he picked one that:
1. Linked to a video,
2. Was written poorly enough to be nearly incomprehensible, and
3. Called 9 fps "moving along pretty well."
Slashdot is jumping the shark. I'm just driving the boat.
I was expecting all sorts of bits coming together in this article, such as image/video processing, adding several images together, object recognition, the Heisenberg uncertainty principle as applied to video images (if it's blurred, you probably can't recognise what it is - but then, it's probably moving fast, so precise identicication takes a lower precedence than getting out of the way), attempts to put all gathered data into perspective, i.e., making a "map" of the outside world, and so on.
... which might actually be an improvement on some people's driving, come to think of it.
What do I get? Hey! I remember thses things from the eighties. You got a little bug with a couple of light sensors underneath, plus a red pen. You also got a thrashing when you covered the most available big flat white surface, which was the kitchen table, with roads (a.k.a. "scribbling").
OK, so they've gotten faster, but it looks like any car you upgrade with this tech will race down the centre marker of any highway you let it loose on
yes, we have no bananas
9 ft/s, eh? Can we have that in furlongs per fortnight please? I am so much more accustomed to the unit.
For the metrically non-challenged:
9ft/s = 2,7432 meter/s
thats = 9,87 km/h or 6,13 miles/h
And for collectors of odd figures and measurements: that speed equals 7769 nautical miles per presidential term.
+++ath0
CogniToy has a game that simulates the programming of these robots. They have a small web based demo Here If you go to their main page you can get more details on the full game
A few years ago when F1 had every driver aid under the sun there was a rumour, urban legend?, circulating that Mclaren had a car that was lapping Silverstone with out a driver.
I was hoping we would see the first car to win a GP without a human. Looks like Schumacher has achieved that now.
it would be easier to crated one for nascar then for 'civilian' use. there are far more variable in everyday use then at NASCAR.
You can write a program to 'predict' what other cars are going to do just as well as a human can. the trick is to get the right response quick enough to use those 'predictions'.
NASCAR driver do things for very specific reasons.
note: I said easier, not easy.
I would wager, that if there was a "robotic NASCAR" it would only be a few years before the AIs could outperform there human counter parts.
Plus it would be cool to see how the cars start to change when there is no human driver in one.
The Kruger Dunning explains most post on
Listen you arrogant,stuck up, whiney bastards.
The site is in imperial measurements.
What is timothy supposed to do, tell them to change there measurement type, then post the story?
The Kruger Dunning explains most post on
> Your use of the word "throughput" reveals that
> you've got the completely wrong model in your head
I think that you have it backwards. Throughput is exactly what we are talking about. There is a reason that the same terminology is used in data transport design and people transport design. The reason being, that is the same thing that we are measuring or talking about.
Throughput, traffic capacity, latency, all apply, with the same basic concepts to both fields of engineering.
Latency is how long it takes to get from one end to the other.
Throughput is how much can go through in a given time.
Just as with data transport, in people or car transport, throughput is measured in "desired-units" transported per unit of time. We are talking about bits per second, bytes per second, or cars/people per hour. The same principles apply. Make the "desired-unit" shorter and throughput goes up. The spacing of bits at 100MHz is shorter than at 10MHz. A 100MHz data transport has higher throughput than a 10MHz one, all else being equal. In data transport the actual speed is nearly fixed so the only options are to add lanes(make a wider bus) or decrease the spacing of the bits. In the people transport field there is more room to maneuver.
A road that is densely packed with narrow inter-car spacing will have high throughput and long latency. The long latency is produced because the short spacing forces the drivers to slow down due to their limited reaction time and brake performance.
The optimum condition for a roadway is when the cars are operating at the road's maximum possible speed and the minimum spacing allowable at that speed. This condition produces both minimum latency (which is very important in people transport) and the maximum throughput possible while not increasing latency. Add one more car and the spacing will be reduced, forcing the speed to be reduced, causing an increase in latency. This is the "people transport" definition of congestion.
In the data trasnport field congestion also arises when conditions force latency up. The data transport might be operating at peak throughput, but there is more data to move than in can handle. Tha data is forced to wait and therefore latency goes up. This is the data transport definition of congestion. As in the people transport case, operating conditions have forced latency to increase.
All this talk about computer driven cars comes from the need or desire to improve(reduce) reaction times and so reduce the spacing requirements for any particular speed. The alternatives, mass transport or car pooling, operate on the same principle. Notice that while I've been talking about car spacing, I've also been using the term "people transport". Moving people and their stuff is what this is all about. Putting more people, or more stuff in a vehicle also reduces the "people spacing" requirements of a road. As such it also reduces congestion and improves throughput. The same effect is produced in the data transport field by compression.
The same effects produced by compression can be seen in both fields. Compression can improve throughput but can also have a negative effect on latency. Data must wait in a buffer to be compressed and then expanded. People must navigate to and then wait at a bus stop for the bus and then do the reverse at the other end. Depending on the situation in both the data and the people transport scenarios throughput can be improved or reduced as can latency.
In the data transport situation, you rarely start using compression until the latency or throughput limitations become intolerable. The same thing happens in the people transport arena. Given a wide open road, good weather, and a good car, I'm driving. Given instead a traffic jam, I'll take the bus and read the paper on the way...
Good judgement comes from experience, and experience comes from bad judgement.
- W. Wriston, former Citibank CEO
No where have I tried to treat it as such.
In that respect, your comparison of data transport and water plumbing is completely off the wall. If anything, data transport is much more closely related to roadway transport than fluid transport via plumbing.
All I have have done is try to explain the design principles, terminology definitions and the similarities between fields of endeavor. You might think that the principles that govern the design of roads used by a few hundred thousand drivers all doing their own thing doesn't relate at all to data trasnport design, but you will be mistaken. The models all break down when you try to delve into details of individual cars and include accidents and such, but the overall model of the road and how it works does not.
If a crash closes a lane, two things happen. For one, the road can now be modelled as one with one less lane. Two, the approaching drivers within visual range of the wreck can be modeled as having grossly increased reaction times.
The effects are right in line with the principles outlined earlier - increased congestion.
After passing the wreck, the drivers reaction times are reduced back to normal and the road expands by one lane and bingo - free flowing traffic in near optimal conditions.
Good judgement comes from experience, and experience comes from bad judgement.
- W. Wriston, former Citibank CEO