Print 20 comment(s) - last by grant2.. on Jul 29 at 6:41 PM

System is designed to be less-intrusive then full-autopilot

The Massachusetts Institute of Technology (MIT) has developed a system called "intelligent co-pilot".  Rather than aiming for fully autonomous artificial intelligence-driven driving, à la Google Inc.'s (GOOGself-driving car project, the new MIT study focuses on a "semi-autonomous" system.

I. Hunting for Safety

According to Sterling Anderson, a PhD student in MIT’s Department of Mechanical Engineering, most current commercially implemented algorithms hunt for static clues in their environment, such as the curb.

This isn't how the human driver functions.  Comments Mr. Anderson, "The problem is, humans don’t think that way.  When you and I drive, [we don’t] choose just one path and obsessively follow it. Typically you and I see a lane or a parking lot, and we say, ‘Here is the field of safe travel, here’s the entire region of the roadway I can use, and I’m not going to worry about remaining on a specific line, as long as I’m safely on the roadway and I avoid collisions."

To mirror that human mind-set the MIT team's algorithm uses so-called "homotopies" -- probable safe zones in the environment.  The environment is triangulated, as the driver is drives to determine if the driver is crossing the border from safety to danger.

When such an event is detected, the car's AI takes over and steers the car around the obstacle, back into a homotopic (safe) zone.

The team has performed 1,200 trials in which they drive normally, but then abruptly head on a collision course with a construction barrel.  Most times the car has been able to avoid the collision.  The few incidents where there was a collision appear to have stemmed from camera failures.

The tests were performed on a course in Saline, Mich., a city in the state's southeast Washtenaw County.

Eaton Corp. intelligent truck technology manager Benjamin Saltsman praises the system for its minimalist approach.  He says that the system uses less computation power and fewer sensors than fully autonomous alternatives from Google and Ford Motor Comp. (F).

Comments Mr. Saltsman, "The implications of [Anderson's] system is it makes it lighter in terms of sensors and computational requirements than what a fully autonomous vehicle would require.  This simplification makes it a lot less costly, and closer in terms of potential implementation."

II. Next -- Using a Smartphone

Mr. Anderson isn't completely satisfied with the system, however, as he fears it could lure beginning drivers to rely on the collision avoidance as a crutch and perform more risky maneuvers.  On the flip side of the coin experienced drivers may be frustrated with the system for overriding dangerous maneuvers.

Still, he's convinced the technology may be eventually fine tuned to be relatively pleasing for the masses and save lives.  

Fresh off a presentation at the Intelligent Vehicles Symposium in Spain, hosted by the Polytechnic School of the University of Alcalá in Madrid, Spain, the team is working to scale down their invention to an even simpler system.

Mr. Anderson and Karl Iagnemma, a principal research scientist in MIT’s Robotic Mobility Group, the other author of the work, will next look to use a dashboard mounted smartphone (which offers a camera, accelerometer, and gyroscope) to perform identical collision detection.

The MIT researchers next want to make the system capable of running
on a cell-phone using only its minimal sensors.

The ongoing research is funded by grants from the United States Army Research Office and the Defense Advanced Research Projects Agency.

Source: MIT

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

One barrel?
By othercents on 7/16/2012 5:04:05 PM , Rating: 2
I can understand avoiding a single car collision where a driver hits a pole or parked car, however avoiding a multi-car crash where there are multiple moving parts with degrees of safety in a decision is much more complex. I wonder what the system will do if you are going 10-15 mph into a car wash?

I say let me drive the car or have everyone not drive. As long as there is human interaction you will have accidents.

RE: One barrel?
By Etsp on 7/16/2012 5:10:01 PM , Rating: 4
There will certainly be instances where this technology will not be able to prevent a collision. Often those instances would be the same as when a human driver would not be able to prevent a collision.

However, I am curious to find out how it would deal with mechanical failure of the vehicle. For example, a blown tire at 75+mph. Or some other sort of mechanical failure that would cause the vehicle to not react normally.

RE: One barrel?
By WalksTheWalk on 7/16/2012 5:36:25 PM , Rating: 3
I'm sure it can prevent all kinds of basic collisions with a simple choice: slow down, turn left, etc. Is it smart enough to know to take the least destructive course of action into account?

For instance, if a car pulls out in front of me and a collision is imminent unless I swerve onto a busy sidewalk, what will it do? Will it take the collision with another car where the likelihood of survival is almost assured but the car will take damage, or will it swerve onto the sidewalk mowing down unprotected bystanders?

RE: One barrel?
By wordsworm on 7/16/2012 9:45:30 PM , Rating: 2
I believe the idea is that the computer can be programmed to make the same decisions for the same emergency situation every time. While the person who is busy responding to a text message looks up at the last second (or not) simply spins the wheel in some random direction in a moment of panic, either mowing down the public or colliding with another vehicle.

That said, I'm still waiting for the day when I can push a button on a car and it'll take me where I want it to go while I take a nap.

RE: One barrel?
By Samus on 7/16/2012 10:36:12 PM , Rating: 2
It's like any safety device: what are the benifits and what are the drawbacks.

Antilock brakes can cause accidents, especially in snow where they will slow you down slower than by manually pumping the pedal slowly (treadgaping the packed snow into the tire instead of creating a pileup against the tread like ABS tends to do when not letting the tire slip enough)

However, ABS saves many lives in high speed scenerios on dry pavement and damp surfaces by distributing the tread area to absorbe equal amounts of energy.

Airbags, especially early implementations, have definately killed people, unfortunately mostly children and the elderly.

However, unarguably, airbags, especially the side-impact variety, have saved more than those that have been killed by them.

Seatbelts are a no brainer and obviously save more than they kill (if there is even a record of someone being killed as the direct result of a seatbelt who would have otherwise survived had they not been wearing one...)

All safety technology has drawbacks, but generally, the benifits outweigh them.

RE: One barrel?
By Etsp on 7/17/2012 12:18:50 AM , Rating: 3
There are a large number of seat-belt related injuries where passengers in the same vehicle weren't hurt badly. Improperly worn seat-belts have caused fatalities as well.

As a counter-example, my aunt was thrown from her van when it rolled, her husband and mother-in-law were crushed inside when the ceiling collapsed. Not wearing a seat-belt saved her life.

However, I am a firm believer that wearing a seat-belt is far safer than going without.

RE: One barrel?
By Gurthang on 7/17/2012 9:40:09 AM , Rating: 1
Yea, that is not a good example like the ones the previous poster made. In your case the death was caused by colapse of the vehicle structure. The seatbelts did their job keeping the occupants in the usually safer location inside the vehicle. Had the structure not collapsed so badily or the location of that person's "exit" from the vehicle resulted in a being thrown into a oncomming traffic or worse we would not be having this conversation.

All saftey equipment has its limits and mostly they are one of cost. Though sometimes technology of the time limits what can be done. Like the refinement of airbag systems, most of the sensors and CPU power that enable all of that control just didn't exist when they were first created. Even seatbelts with the modern pretensioning and seat designs to keep you in the right spot and the belt at it most effective beat the heck out of those old school lap only belts that came first.

These technologies have a place and even if we don't want to admit it very few of us are as good a driver as we often claim to be. I know I would not mind something like this that might prevent the car from hitting something in my blind spot or prevent me from rear-ending someone because they suddenly hit the breaks while I was focused on changing lanes in busy traffic.

One of my favorite car improvement ideas has been to take the rear camer idea and use a funny mirror on the antenna mast and some software to correct the distortions so that you see not just the rear but the whole blind "zone" almose as if you were flying above the car. Though I think that a HUD and some simple lidar/sonar/radar which displays nearby car level obstructions with a little clean-up might work as well if not better.

RE: One barrel?
By inperfectdarkness on 7/17/2012 1:10:19 AM , Rating: 2
I'm just amused that this guy is named "Missster Anderson".

RE: One barrel?
By kwrzesien on 7/17/2012 2:53:59 PM , Rating: 2
I say require ALL vehicles to be automated on the freeway. I for one welcome our new smartphone/Google/driving overlords.

RE: One barrel?
By Ammohunt on 7/17/2012 2:46:01 PM , Rating: 2
I agree! this is a complex solution to an otherwise simple problem that being poor drivers. All that needs to happen is make 2 or more years of driving school mandatory and the driving test more stringent in order to weed out people that should not be driving.

By Solandri on 7/17/2012 1:11:07 AM , Rating: 5
What if the computer incorrectly identifies a scenario as safe because the programmer hasn't thought of it (e.g. piano being lifted on crane falls from the sky), and overrides the driver's frantic efforts to steer out of the way, thus insuring the piano hits the car?

Ostensibly the human is in control unless the computer detects a situation which is dangerous. But in reality the computer is always in control since it decides when to override the human, not the other way around. In the 1980s or 1990s, this led to a spate of runway overrun incidents by Airbus planes.

The planes were landing onto rain-slickened runways and refusing to engage the thrust reversers. Obviously a thrust reverser deploying during flight is a catastrophic event, so Airbus programmed the plane to be absolutely certain it was on the ground before deploying them. Unfortunately when landing on wet runways, sometimes the wheels hydroplaned and didn't start spinning. The computer interpreted this as an indication that the plane was still flying and thus refused to deploy reverse thrust. Wheel brakes don't work well when you're hydroplaning, so the planes went off the runway leading to many fatalities. IIRC the wheel gear weight sensor can now override the wheel's spin sensor - if the gear is compressed by the weight of the plane, the computer says it's on the ground even if the wheels aren't spinning.

But you can see the problem here - a scenario not envisioned by the programmer leads the computer to misinterpret a situation and prevent the human from taking corrective action. I think I prefer Boeing's approach to this. On Boeing planes, the computer can override the human, but if the human really pushes the controls against the computer's wishes, it relinquishes control back to the human. Boeing can do this because their yoke and thrust levers provide force feedback (you have to exert more force to move the controls against the computer or aerodynamics). The Airbus controls provide no force feedback - they're basically just position sensors for the pilots to send input to the computer, not the other way around.

So the way I'd envision the piano scenario playing out is the driver sees the piano and swerves the car to avoid it. The computer interprets the swerve as a dangerous maneuver and causes the car to continue straight. The driver feels this as the steering wheel resisting his attempt to turn it. But if he continues and puts more muscle into it to force the steering wheel to turn, the computer interprets this as an override and gives control of the car back to the driver.

(Then some patent troll will come along and patent this idea which is already in use on planes, because it's totally brand new and innovative when used on cars. *rolls eyes*)

Ah yes
By Dr of crap on 7/17/2012 12:23:13 PM , Rating: 2
Don't you just love it when the "learned" people label something we peasants do every day with some stupid label.

Really - homotopic?
My homoauto doesn't really care about my homotopic issues, because I control them very well, thank you very much!

RE: Ah yes
By ppardee on 7/19/2012 2:41:40 PM , Rating: 2
Homotopic is probably the best word for what it's doing... Homotopic means of the same height, essentially. There are more complex connotations, but it doesn't mean 'safe'...

What we peasants do on a day-to-day basis is simply astonishing. Trying to make a machine do it requires defining what we think we do intuitively. You will instinctively not continue to walk if there is a huge change in topography in your path (if you notice).

Preventing crashes
By Zhukov on 7/22/2012 7:59:56 AM , Rating: 2
Why can't MIT people like Anderson just make a car that will obey the speed limit? That would save 10,000 lives, countless head and neck injuries, and $30 billion in medical and auto damage costs each year in the US. It would be simple and cheap to limit a cars speed to the posted limit. In that case all cars would be going about the same speed on on highways and not speeding through traffic lights, etc. There would be a lot less road rage too. They keep doing more and more research on cars that will drive themselves, but these cars are still in the future. Automatic speed control can be done now and cheaply. Most people would see the advantage of a car that can't exceed the speed limit - no risk of a speeding ticket, fewer tailgaters, increased safety, and by the way, it's already illegal to exceed the legal limit, so nobody's rights would be compromized.

RE: Preventing crashes
By grant2 on 7/29/2012 6:41:31 PM , Rating: 2
Do you actually know anyone who always obeys the speed limit??

I think my late grandmother used to, back in the 80s...

By nofear4COMment on 7/16/2012 7:16:28 PM , Rating: 2
He should be pay a visit to IBM for advice, IBM has spend billion of dollars for super computer to master this scenario.

How about this situation
By Kurz on 7/17/2012 10:12:43 AM , Rating: 2
You have two choices either going under a Semi trailer where you'll lose your head. Or crashing into another car?
Which is safer?

I bet the system will fail to realize the situation and force you to take the former.

By drycrust3 on 7/17/2012 11:46:17 AM , Rating: 2
When such an event is detected, the car's AI takes over and steers the car around the obstacle, back into a homotopic (safe) zone.

I think the last thing I'd want in a potential accident situation is a machine taking control and deciding what it thinks is the best solution.

By Prescott_666 on 7/17/2012 12:51:54 PM , Rating: 2
When will you be able to drive home after having a few, if you have the driver assist installed and turned on?

To be an intelligent co-pilot
By YashBudini on 7/17/2012 11:20:31 PM , Rating: 2
It would have to determine if the driver is on the phone and then it would keep repeating in an ever increasing volume, "Shut up and drive."

"Vista runs on Atom ... It's just no one uses it". -- Intel CEO Paul Otellini

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki