In this encore edition of the podcast, we revisit a 2018 interview with Beau Woods of The Atlantic Council from episode 89. Beau and I talk about the perils of autonomous driving software and whether automakers and regulators are rushing too quickly introduce autonomous features without adequately understanding the risks they pose.
Elon Musk has been busy on social media in recent week. Among his usual rounds of SpaceX fanboy videos, memes and Dogecoin boosterism, Musk has regaled his massive social media following with the promises of a software update for Tesla vehicles: so-called FSD v9, or “version 9” of the company’s Fully Self Driving software.
The software was released over the weekend and pushed out to Tesla vehicles soon thereafter. According to media reports and posts by Tesla users, its pretty impressive: allowing Teslas to navigate city streets – not just highways – pick their way across busy intersections and follow GPS directions to choose forks without driver intervention. The on-board “Tesla Vision” also got a big update, identifying whether cars on the surrounding roadways are slowing, accelerating or stopped…and more!
Beta Software Behind The Wheel
As to whether the “beta” software is safe and reliable now that it has been pushed out to millions of vehicles navigating U.S. roadways? That’s another question entirely – and one without a clear answer. Musk himself has admitted that automated driving is a much more complex problem than he initially estimated.
And his company seems eager to cover its butt. The company’s announcement of the FSD v9 release said less about the autonomous driving features in the update than about the abundance of caution Tesla drivers should use.
Autopilot Accidents Pile Up
The release – and the hype surrounding it – come at a curious time. As accidents linked to Tesla and its “Autopilot” technology pile up, the company finds itself in the cross hairs of the National Highway Traffic Safety Administration (NHTSA). In June, for example, NHTSA instructed automakers to begin reporting and tracking crashes involving cars and trucks that use advanced driver-assistance technology such as Tesla’s Autopilot and General Motors’ Super Cruise.
Critics charge that merely calling the feature “Fully Self Driving” encourages drivers to think of it as just that – resulting in accidents. In response, Tesla has publicly walked back Musk’s boosterism with words of caution about the fallibility of the software.
Which brings us to the obvious question: if the software and the features aren’t safe – and could even be deadly, why are they being pushed out to millions of vehicles? And, beyond Musk’s casual admission that autonomous driving is harder than he expected, how realistic is the vision of fully autonomous vehicles, anyway?
We’re Close To Autonomous Driving. Or Are We?
Those are questions we posed to Beau Woods back in 2018. Back then, Woods was a cyber safety innovation fellow with the Atlantic Council. He is now a Senior Advisor at the Cybersecurity and Infrastructure Security Agency, where he assists the COVID Task Force in defending the healthcare supply chain against cybersecurity threats.
In this archive podcast conversation from episode 89, Beau and I talk about a fatal accident in which an autonomous vehicle operated by ride sharing firm Uber struck and killed 49-year-old Elaine Herzberg as she walked her bicycle across a Tempe, Arizona street.
In this conversation, Beau and I talk about the need for automakers to be thoughtful about how they develop and deploy autonomous driving systems – and the tendency of autonomous driving boosters – like Musk – to underestimate the complexity of the challenge.
To start out, I asked Beau to reflect on the implications of the recent incident in Tempe involving the death caused by an autonomous Uber vehicle.
As always, you can check our full conversation in our latest Security Ledger podcast at Blubrry. You can also listen to it on iTunes and check us out on SoundCloud, Stitcher, Radio Public and more. Also: if you enjoy this podcast, consider signing up to receive it in your email. Just point your web browser to securityledger.com/subscribe to get notified whenever a new podcast is posted.