Auto pilot: Above is a version of Junior, Stanford's autonomous car. While this car can chart a course without a driver or remote assistance, it’s not yet equipped with all the sensors necessary for it to compete in the Urban Challenge, a race in which autonomous vehicles navigate city streets. Below is the laptop exhibiting the course instructions for Junior. Data from the laptop is sent to Junior's computers, which guide the automated driving system.
Kate Greene, Technology Review.
Junior the autonomous car passes an initial round of tests in preparation for DARPA's Urban Challenge, slated for later this year.
Yesterday, a Volkswagen Passat drove around a parking lot in Mountain View, CA, made three-point turns, and followed the rules at a four-way stop--all without human intervention. The computer-controlled car is named Junior, and it's Stanford University's official entry in the DARPA Urban Challenge, a race in which an autonomous car must navigate city streets, obey traffic laws, avoid obstructions, and, crucially, drive well among other cars in traffic. This test run is Junior's first public appearance, designed to let DARPA (the Defense Advanced Research Projects Agency) test the car and determine if it proceeds to the next round in the Urban Challenge.
The motivation for the Urban Challenge is to build a better car. "As we all know, cars are unsafe," says Sebastian Thrun, the team leader and a professor of computer science at Stanford. Car accidents kill roughly 42,000 people a year in America and about a million people worldwide, he says. In addition, cars are inefficient, causing traffic jams and requiring people to consistently focus on the road during long commutes. The Stanford researchers' goal is to design a car that can drive itself, conceivably making roads safer and giving people back their time. "The idea of a self-driving car, in my opinion, will change society," says Thrun.
Two years ago, the Stanford team, using a computerized car named Stanley, won DARPA's Grand Challenge, an autonomous-vehicle race in the desert. That car had a number of GPS sensors and lasers, a camera, and other equipment to help it make its way through the course. Junior is based on the same fundamental technology, says Thrun, but with some crucial improvements.
Junior uses the same kind of laser perception as Stanley, but with longer range. The new car has a total of eight LIDAR systems that emit beams of light and detect reflections to determine the distance of other objects. One system is mounted on the front of Junior's roof and has a range of about 100 meters--many times that of Stanley. Another LIDAR system points at the ground and constantly keeps track of the road and reflective lane markers. A third system constantly takes a 360-degree image of its surroundings. All this data is process by two Intel quad-core machines running at 2.3 gigahertz, and the pertinent information is relayed to the driving systems, which guide the car.
Junior is also equipped with a precise location system that include GPS and other sensors that measure the revolution of the wheels and the direction the car is moving in. Together, these sensors allow Junior to pinpoint its location to within 30 centimeters.
Importantly, says Thrun, Junior has a lot more "intelligence" than Stanley so that the computer can deal with intersections and traffic. Such tasks simply weren't a part of the previous race, which basically involved driving down a curvy desert road. This intelligence comes in the form of about 500 different probabilistic algorithms that process all the environmental information collected by the sensors and make the decision that is most likely to be the best. Thrun says that these decisions are made in less than 300 milliseconds, which is sufficient for slowing down or changing lanes if a car in another lane tries to merge into Junior's. "In the last race, you basically only had to decide whether to speed up or slow down," says Thrun, "but this time there are discrete decisions on top of that."
In Thursday's test, Junior successfully completed all the tasks DARPA assigned. The first was a safety test that ensured that the team could remotely stop the car from a speed of 20 miles per hour. The other tasks included navigating within a lane, stopping at stop signs, making U-turns, avoiding obstacles, and following driving instructions that DARPA provides.
Junior will move on to another test run in October and, if all goes well, eventually compete in the final round of the Urban Challenge on November 3.
Researchers figure out how to hijack sensor communications.
On Saturday, 11 autonomous vehicles designed by the best roboticists in the world will compete for a $2 million prize in the Urban Challenge.
Just to make the image more clear, the car pictured was above was not the car tested. The highly logo'd car in the picture is the current "backup" car for Junior which is lightly logo'd so that it's street legal to drive to test sites etc. The logo'd car was used to give the crowd free autonomous rides using only GPS, I took some rides just to get in its air conditioning, but it is still "cool" in more ways than one. The official Junior has much more perception hardware on top.
Eventually the cars will "swap" roles with the "show" car becoming the official car and the current "official" car becoming the backup. The reason is simple -- put the wear and tear on the test car, but use it only as a backup on race day.
I'm impressed that they added new probabilistic algorithms and have as much processing power as they do, but I'm still waiting to see if the cars can perform on the horribly maintained roads that we have. Will they continue to drive on lanes even though the reflective markers have been etched off and the lanes have been redirected by cones, or nothing at all? Will they obey the signal person, or the police officer? Will they yield the right of way on a one-lane street, or will they continue if it was their turn, or will they do nothing - just sit there and wait while the other driver gets mad, or the other computer waits too?
And then there's the other drivers, who will all do their best to f*** with it's head. Computer. Whatever.
;-P
IMHO, I think what Sebastian Thrun has keyed in on is that the pieces are now in place to convert the conceptually very hard AI problem of autonomous driving into a step by step engineering program. First autonomous driving across empty desert road was solved because better sensors (Lidar, GPS, odemetry, and cameras) met maturing software (Sigma point kalman filters, monte-carlo type panning etc).
Look at the new sensors on this car -- again, better 360 lidar and side views are meshing with better planners supported by now dual quad core processing -- much more processing at less energy.
Other hard problems navigation problems can be solved by prior data gathering rather than perception along the lines of google street view.
As you point out, it doesn't end there. Road hazards etc etc. But after the urban challenge, we already have a huge boost for ordinary cars (once the sensor and processor price drops) -- it can keep your car in it's lane when you are not actively steering and can keep you from smashing into looming objects. That's enough economic incentive to further fund the solution of road hazards on to completely autonomous driving.
I think the obvious answer to your question about what an automated car would do when it encountered a policeman or traffic cones is that the police and construction crews would be equipped with appropriate reflective devices (cones, tape on road, reflector on pole, etc) that the car would recognize. I envision the police & signal people remotely controlling a device from a position safely off the road, simply ensuring that traffic continues to flow around the hazard.
If you think about it, dealing with computers that have a list of instructions is far safer than dealing with the drunk drivers that need to avoid a stalled car at 2am.
Cars driving closer to one another will help eliminate highway congestion and humans cannot do this safely. Increasing the speed limit would have no effect because of the two second following distance rule. You still have 1 car passing point x every two seconds no matter what speed they drive at - unless they drive closer to one another. Bring on robotic cars. How would a robotic car following at short distance at high speed react if a bicycle fell of the back of the car in front? The robotic car would sense the impending disaster before the bike even hit the ground. It would need to communicate to the cars next to it and behind it so as to create a gap. This gap creation would start happening before the bike had even hit the ground. If there was no communication between the cars, the emegency reaction, compounded by problem - movement - reaction time delays could be beyond the abilities of the car.
When a car can drive in traffic, hold a conversation with a passenger, answer a cell phone, eat a donut, and listen to music, all at the same time, I will be willing to ride in it. Until then...
But, of course, many human drivers can't do all that successfully.
just think what it will do to reduce premiums once the technology gets ironed out. I pay 1,500 for 3 cars every 6 months. I'm hoping technology like this can cut it in 1/2 if not 1/3. All the drivers over 60 will reduce their accident rate by 99% i'm sure. And fashionable women who do the lipstick in the mirror while answering a blackberry they have for no apparent reason will be thankful as well.
Great society changing technology. The only problem i see with it is that i'll be expected to log onto my email via my laptop while on my way to work! More work..less down time. But i will be thankful..because i dislike driving that much!
( Please forward contents of this mail to proper authority for decision)
Dear Sir,
We do not know whether you will believe it or not.
But it’s our privilege to introduce a unique solution for “Driverless vehicle” technology.
Please permit us to put before you for introduction a developed and tested new breed “DRIVERLESS VEHICLE” technology on our newly created “Foot Print” technology, with out the use of computer, Radar, night vision, GPS and or any other computer based technology.
Our “Foot Print” technology has claimed the followings:
1. Eliminating the use of computer, Radar, GPS, night vision or any computer base technology in “Driverless vehicle”.
2. Technology can convert existing automobiles (Light or heavy) into New breed “Driverless vehicle” on as and where basis.
3. The cost of New Breed “Driverless vehicle” based on our ” foot print” shall be drastically low( More than 50%) compare to the “Driverless vehicle” base on existing and proposed technology.
We are ready with the working model to demonstrate the working of the technology and new breed “Driverless Vehicle” for indoor and long distance outdoor application.
We are looking for a technology buyer /sponsor
We shall feel honored to answer your quarries.
From
(Rajendra Singh Chauhan)
B.E.(Mechanical)
M/s Design Engineers
India
Mob: 919893356526
e-mail:chauhan042000@yahoo.co.in
(Please check attachment for more details)
There shall be no harm to try our achievement
Please check out the pemmPOD which can navigate with a guidePOD at
http://dancrissco.wordpress.com/category/sustainable-cities/pemmpod/
It would work in small self sustainable cities with planned stripe painted pemmPOD lanes.
The concept is to have this as a service rather than a product with assembling modular components
Manufacturing in the United States is in trouble. That's bad news not just for the country's economy but for the future of innovation.
This document is part of the “How-To Guide for Most Common Measurements” centralized resource portal. This tutorial provides a detailed guide for measurement and device considerations to take temperature measurements using thermocouples. Get an introduction to thermocouples, which are inexpensive sensing devices widely used with PC-based data acquisition systems. Also review some specific thermocouple examples and learn how thermocouples work and ways to integrate them into a data acquisition measurement system.
View full PDF >
Listen to story > 
On Twitter
Become a Fan on Facebook
Subscribe to the Feed
Phineas
127 Comments
Diesel Power
I see from Junior's fact sheet that it's a TDI diesel. 43/35 mpg on an automatic transmission. Diesel powered cars account for nearly half the traffic in the EU, less than three percent in the States. You might have to wait in line for one of these gems.
http://cs.stanford.edu/group/roadrunner/pdfs/final_factsheet_junior.pdf
Reply