I, too, am left wondering over the paradox of the brake supposedly being powerful enough to hold back any amount of torque the engine and transmission can apply - unless, in the name of anti-skid braking or traction control, the system has the ability to override the driver and release or limit the braking torque.
Furthermore, those explanations of the floor-mat recall that I have checked all say that the mats were suspected of having interfered with the accelerator, not the brake. Toyota subsequently performed a second recall, to correct problems with the accelerator pedal. These both imply that unintended acceleration was at least a significant causative factor in the accidents - unless Toyota and the NHTSA were simply acting in order to be seen doing something.
With regard to proof, the article covers the issue in some detail: after a far-from-exhastive series of investigations, there have already been found so many ways for the software to get into unintended states that only a vanishingly small number of the vast range of possible causes have been considered. Asking for proof of an error is the wrong question to ask: the burden of proof clearly rests with the manufacturer to show that their systems are safe - and Toyota cannot do that with this software.
That Toyota replaced floor-mats is certainly not evidence for the correctness of their software, and nor is the non-appearance of a software patch. The article quotes a concerned Toyota engineer whose statements indicate a culture of denial within the company.
Furthermore, those explanations of the floor-mat recall that I have checked all say that the mats were suspected of having interfered with the accelerator, not the brake. Toyota subsequently performed a second recall, to correct problems with the accelerator pedal. These both imply that unintended acceleration was at least a significant causative factor in the accidents - unless Toyota and the NHTSA were simply acting in order to be seen doing something.
With regard to proof, the article covers the issue in some detail: after a far-from-exhastive series of investigations, there have already been found so many ways for the software to get into unintended states that only a vanishingly small number of the vast range of possible causes have been considered. Asking for proof of an error is the wrong question to ask: the burden of proof clearly rests with the manufacturer to show that their systems are safe - and Toyota cannot do that with this software.
That Toyota replaced floor-mats is certainly not evidence for the correctness of their software, and nor is the non-appearance of a software patch. The article quotes a concerned Toyota engineer whose statements indicate a culture of denial within the company.