When it comes to sudden acceleration, automakers claim absence of a detectable fault proves absence of design defect. Not true, said a panel of electronics experts who gathered Tuesday at the National Press Club in Washington, D.C., to demystify some of the electronics issues they say are likely responsible for Toyota – and other automakers — sudden acceleration incidents.“Thirty years of empirical evidence overwhelmingly points to sudden acceleration being caused by electronic system faults that are undetectable by inspection or testing,” said Keith Armstrong, an electronics and EMI expert and Interference Technology editorial board member. “Like Newton and gravity, we should deduce electronic failures from the actual vehicle performance.”The contentious issue of electronics failures is at the heart of an investigation by Congress and the National Highway Traffic Safety Administration (NHTSA) into the more than 5 million vehicles recalled by Toyota following growing reports of sudden acceleration, said Joan Claybrook, a former NHTSA administrator and moderator of Tuesday’s panel. Toyota has pointed to floormats and sticky gas pedals as possible causes of the problem.Toyota has sworn under oath to the Congress that their vehicles do not have electronics problems, Claybrook said, and the automaker is liable for criminal penalty should it be found that the company lied to investigators. “Since it’s clear that Toyota will continue to deny there’s a problem, we in government must turn to experts who can dissect this issue for us,” she said. Vehicle throttles used to be directly controlled with mechanics, but now control is indirect via electronics, Armstrong said. The problem is electronics have weaknesses and can go wrong in many ways. Similar to rebooting a computer, where no defect can be found after a restart, electronic malfunctions in motor vehicles often leave no trace.“It is easier for people to blame drivers, floormats, sticky pedals, whatever diverts attention away from the electronics,” Armstrong said. Automakers claim they conduct comprehensive EMI testing, but their tests don’t cover most real-life EMI, Armstrong said. “They don’t simulate typical faults to verify their back-up or fail-safe measures work during the electronic interference testing.” Besides, no practical amount of testing can ever be sufficient in the first place, he said. Based on rate of occurrence, to detect sudden acceleration in just one model would require testing 36 vehicles, 24 hours a day, seven days a week, for 10 years. To test a single vehicle, it would have to be driven for about 200 million miles. Automakers also claim their electronics are totally shielded against EMI but, in fact, this is not achievable, Armstrong said. “Even military can’t do this,” he said. “You can make things very good but you can’t make them totally shielded.”Some auto manufacturers use unshielded plastic connectors for their engine control modules, and Toyota gas pedal and throttle sensors use only partially shielded plastic housings, which Armstrong compared to a leaky umbrella. “If you say you’ve got a good umbrella but there’s a hole in it, it’s not really that good,” he said.In disputing automakers claims that their software is “bug-free,” Armstrong pointed to research by Carnegie Mellon University, which says the high-quality code used in the Space Shuttle has about 1 latent bug per 10,000 lines of code. A typical modern car, Armstrong said, has more than 20 million lines of code, which means at least 2,000 undiscovered bugs should be expected in every modern vehicle.The fault codes and black box recorders in cars rely on the very electronics that they’re supposed to be monitoring, Armstrong said, whereas in other industries, such as aviation, the black box recorders are totally independent systems. In order to make electronics safer, Congress should mandate that NHTSA require automakers to comply with the standards for the safe design of the drive-train electronic systems, Armstrong said. “NHTSA must obtain all the drive-train electronic systems and components, their safety specifications, software source codes, hardware design schematics, material, production and quality specifications, test specifications and all the results, and then NHTSA should coordinate expert teams working in parallel to systematically classify the fault potential mechanisms and their context of occurrence,” he said. “They should evaluate the data and use it to monitor and improve safety across the industry.”Contact Sarah Long at [email protected] or 484-688-0300, ext. 24.