ADAS and Autonomous vehicles
Companies with good intentions are shuttered. This, too, is our current reality. As of this writing, there have already been near-misses and several fatalities involving new ADAS capabilities and autonomous vehicles.
Does this mean we’ll halt development and focus on making human drivers safer?
Of course not. Just like we didn’t stop production of airplanes after the first crash or, for that matter, automobiles. But today’s environment is a far cry from the environment of the early 20th century. Technology is advancing at an incredible pace. Social media spreads the word of a product flaw across the globe before the media can break the story. And lawsuits are commonplace.
In the 21st century, a single defect can destroy a company. We can no longer give lip service to phrases like “zero defects.” We either live them or wait for the inevitable disaster that will put us on the streets.
We also cannot afford to ignore the public penchant to try new technology. But to accept the challenge is to accept risks far greater than any we’ve dealt with before. Because when an ADAS capable vehicle or a driverless car is involved in a fatality, there’s no more user error or driver to blame. All fault rests with the company whose logo is on the hood. And the suppliers who support that logo.
Our only option, then, is too ensure that the defects—mechanical, electrical, software—are captured early in the design phase. Failure modes are captured both upstream and downstream of the piece of the program that we’re responsible for. Notification of changes are broadcast to global teams within seconds of the final keystroke. Lessons learned are captured in real-time and applied to the current production model and carried over automatically to future models.
In a world where the road will be filled with semi and fully driverless cars, we need a risk management solution that will balance the acceleration of innovation with product quality and safety. And it’s not enough to keep up, we must stay ahead of the technology.