Science Panel Continues Examination of Unintended Acceleration

Science Panel Continues Examination of Unintended Acceleration

By Bill Visnic September 14, 2010

A National Academy of Sciences committee appointed by the National Highway Traffic Safety Administration to research the probable causes of and remedies for a surge in reports and accidents involving unintended acceleration met in Dearborn, MI, Monday for a third information-gathering meeting since the NHTSA was instructed by Congress to form the panel earlier this year.

The latest informational meeting was highlighted by a presentation by Clemson University electronics professor Todd Hubing's research into potential electronic causes for unintended acceleration.

Hubing said that while human error remains the dominant reason for automotive accidents, at least some accidents can be attributed to electronic failures. And, Hubing added, as more control of the vehicle is given over to advanced active and passive control systems, the proportion of accidents caused by electronic problems can be expected to increase. 

Hubing said pinning down potential electronic origins for unintended acceleration remains elusive because the design of contemporary vehicle electromechanical systems - such as the electronically controlled throttles often cited as a probable cause of unintended acceleration - relies on analogue sensors whose accuracy cannot be validated. Such is also the case for "undefined software" and hardware such as microcontrollers, Hubing added.

Hubing's team is studying the electronic controls and software of ten different light vehicles from a variety of manufacturers and is actively testing two models.

With a definitive link between electronic systems malfunction and unintended acceleration yet to be determined, Hubing presented several short-term recommendations to the NAS committee. The recommendations are similar to initiatives proposed by Congress in the Motor Vehicle Safety Act of 2010, which after revisions by both the U.S. House and Senate, currently is awaiting a vote on Capitol Hill.

Hubing suggested:

> All vehicles should be installed with a brake-override system designed to disengage the throttle in the event the brake pedal is being depressed at the same time. Most manufacturers already have these systems in production or are quickly adopting them.
> Drivers should have a way to override the engine control module that would cut power to the electronically controlled throttle. Many current ignition systems are not designed to fully disengage power to the engine's electronic controls.
> Redundancy for the hardware and software controlling the engine throttle.

In the longer term, Hubing suggested the auto industry would be well-served to adopt an "open" standard for design of electronic architectures and software, such as the recent developments of FlexRay and Autosar open-source electronics architectures. He acknowledged that the FlexRay and Autosar initiatives have met with resistance from the competitive auto industry for the perception the open standards may advantage certain automakers, so open-source development should be from an independent source.

Hubing's team also indentified deficiencies in the NHTSA Office of Defect Investigation database that is the repository of most recorded unintended-acceleration reports and accident-investigation data. He said it would be helpful to future study of the unintended-acceleration phenomena if the ODI database consistently recorded VIN information, offered a keyword search and included description of the location and weather conditions when the reported event occurred.

Bill Visnic:  is an AutoObserver Senior Editor. Follow @AutoObserver on Twitter