Japanese  
  Home  |  Top News  |  Most Popular  |  Video  |  Multimedia  |  News Feeds  |  Feedback
  Medicine  |  Nature & Earth  |  Biology  |  Technology & Engineering  |  Space & Planetary  |  Psychology  |  Physics & Chemistry  |  Economics  |  Archaeology
Top > Technology & Engineering > Carnegie Mellon Methods Keep Bugs… >
Carnegie Mellon Methods Keep Bugs Out of Software for Self-driving Cars

Published: June 22, 2011.
By Carnegie Mellon University
http://www.cmu.edu

PITTSBURGH—Driver assistance technologies, such as adaptive cruise control and automatic braking, promise to someday ease traffic on crowded routes and prevent accidents. Proving that these automated systems will work as intended is a daunting task, but computer scientists at Carnegie Mellon University have now demonstrated it is possible to verify the safety of these highly complex systems.

To do so, the researchers first developed a model of a distributed car control system in which computers and sensors in each car combine to control acceleration, braking and lane changes, as well as entering and exiting the highway. They then used mathematical methods to formally verify that the system design would keep cars from crashing into each other.

"The system we created is in many ways one of the most complicated cyber-physical systems that has ever been fully verified formally," said Andre Platzer, an assistant professor of computer science. He and his collaborators, Ph.D. students Sarah M. Loos and Ligia Nistor, will present their findings at the International Symposium on Formal Methods, June 22 at the University of Limerick, Ireland.

"Auto accidents cost society billions of dollars and too many lives, so automated systems that could increase both the safety and efficiency of our roads only make sense," Platzer said. "It would be foolish to move to such a system, however, unless we can be certain that it won't create problems of its own. The dynamics of these systems have been beyond the scope of previous formal verification techniques, but we've had success with a modular approach to detecting design errors in them."

Formal verification methods are routinely used to find bugs in computer circuitry and software; Platzer is a leader in developing new techniques to verify complex computer-controlled devices such as aircraft collision avoidance systems and robotic surgery devices, known collectively as cyber-physical systems, or hybrid systems. He also is a member of the Computational Modeling and Analysis of Complex Systems (CMACS) center, a CMU-based initiative sponsored by the National Science Foundation to apply verification techniques to a variety of complex biological or physical systems.

Using these formal methods to either find errors in automated vehicle control or prove they are safe is particularly challenging, Platzer said. Like other cyber-physical systems, they must take into account both physical laws and the capabilities of the system's hardware and software. But vehicle control systems add another layer of complexity because they are distributed systems — that is, no single computer is ultimately in control, but rather each vehicle makes decisions in concert with other vehicles sharing the same road.

Platzer, Loos and Nistor showed that they could verify the safety of their adaptive cruise control system by breaking the problem into modular pieces and organizing the pieces in a hierarchy. The smallest piece consists of just two cars in a single lane. Building on that, they were able to prove that the system is safe for a single lane with an arbitrary number of cars, and ultimately for a highway with an arbitrary number of lanes. Likewise, they were able to show that cars could safely merge in and out of a single lane and then extended it to prove that cars could safely merge across a multi-lane highway.

Platzer cautioned that this proof has a major limitation — it only applies to straight highway. Addressing the problem of curved lanes, sensory inaccuracy and time synchronization are among the issues that will be a focus of future work. The methods the Carnegie Mellon researchers developed can, however, be generalized to other system designs or to variations in car dynamics.

"Any implementation of a distributed car control system would be more complicated than the model we developed," Platzer said. "But now at least we know that these future systems aren't so complex that we can't verify their safety."




Translate this page: Chinese French German Italian Japanese Korean Portuguese Russian Spanish


 
All comments are reviewed before being posted. We cannot accept messages that refer a product, or web site.If you are looking for a response to a question please use our another feedback page.
Related »

Wang 
1/24/13 

Motion Control Keeps Electric Car's 4 Wheels - And 4 Motors - on the Road
By Ohio State University
Vehicle 
5/14/10 
The Future of the Electric Car, Analyzed from Within the University
By Universidad Carlos III de Madrid - Oficina de Información Científica
The green revolution for cars is just around the corner.  The electric car is part of the almost immediate future, although certainly challenges first have to be met, such …
Car 
7/28/14 

Students Successful in Electric Car World Record Attempt
By University of New South Wales
Sensors 
3/2/11 

Scientists Steer Car with the Power of Thought
By Freie Universität Berlin
Whilst 
7/29/11 
Put the Brakes on Using Your Brain Power
By Institute of Physics
German researchers have used drivers' brain signals, for the first time, to assist in braking, providing much quicker reaction times and a potential solution to the thousands of car …
Battery 
4/19/10 
Electromobility: Fast-tracking Innovation
By Fraunhofer-Gesellschaft
The goal is to help companies speed up their pace of innovation. Researchers will proudly display their initial results at the 2010 Hannover Messe (April 19 - 23, Hall …
Car 
8/21/13 

KAIST Unveils Foldable Micro Electric Car, Armadillo-T
By The Korea Advanced Institute of Science and Technology (KAIST)
Car 
6/2/14 
Laser Device Can Detect Alcohol in Cars, Say Authors in Journal of Applied Remote Sensing
By SPIE--International Society for Optics and Photonics
BELLINGHAM, Washington, USA — A new open-access article in the Journal of Applied Remote Sensing is garnering attention for research that could aid in the campaign to prevent drunk …
Condition 
3/12/11 

Hannover Messe: Smart Materials for High-tech Products
By Fraunhofer-Gesellschaft
Nsf 
6/24/14 

Demonstrating a Driverless Future
By National Science Foundation
More » 
 
ScienceNewsline  |  About  |  Privacy Policy  |  Feedback  |  Mobile  |  Japanese Edition
The selection and placement of stories are determined automatically by a computer program. All contents are copyright of their owners except U.S. Government works. U.S. Government works are assumed to be in the public domain unless otherwise noted. Everything else copyright ScienceNewsline.