You Scared, Bro?

Tuesday, July 14, 2020

#Cybersecurity    #Functional Safety    #Automotive SPICE    #Quality

Maybe Your Autonomous Car Should Ease Your Fears

In a recent survey by Myplanet of various technologies, “autonomous driving” came in as the most uncomfortable of the thirty-five technologies at 66.8% of the Americans surveyed. To put that in perspective, one of the technologies near the middle of the pack was “surgical robot” at 42% negative, which translates into “I’d rather your ‘bot cuts me open than have it drive me to the corner store.” As summarized well by Jason Cottrell, Myplanet CEO, “Customers have made up their minds about autonomous driving and it’s skewed heavily to the negative.” Other studies, in fact, corroborate that level of fear (e.g. AAA’s study suggested 71%), and suggest the cause of that anxiety varies. Some potential customers understand that although robots are intended to erase human error, they are programmed by imperfect humans where software has upwards of fifty (50) bugs per thousand (1,000) lines of code, i.e. as many as 25M bugs in an autonomous vehicle. Also, some are aware that malware infections have risen every year for the past decade and are likely to exceed 1B in 2020, especially since cybersecurity attacks are up 600% during the coronavirus pandemic. And some potential customers are simply afraid to relinquish perceived control as evidenced by 2.6% of the population having a “fear of flying”. Regardless of Functional Safety, Cybersecurity or unfounded root causes, this cumulative fear far exceeds other commonly known fears such as “fear of the dark” (11%), “fear of heights” (7.5%) or “fear of spiders” (3.5%).

So why is so much money being invested in autonomous driving? Maybe Amazon’s impeding $1B purchase of the start-up Zoox makes good business sense considering packages don’t get nervous while being delivered, but it seems like General Motors’s $19B gamble on Cruise should consider this mass trepidation. As the former Vice President of Advanced User Experience Design at Mercedes said as early as 2014, “…designing to gain trust [in an autonomous vehicle] is an amazingly interesting design challenge – far different from [typical] technical problems.” Are there ways to mitigate the fears of the marketplace?

Interestingly, there may be.

Solutions To Consider

 Possibly the most interesting solution to mitigate fear is research conducted by Toyota and SRI International. The idea is to use artificial intelligence to engage with the drivers and alleviate negative emotions, thereby proactively reducing the risk of accidents while offering a more personalized experience. “If you appear to be nervous while driving or hesitant in your maneuvers,” says Amir Tamrakar, a Senior Technical Manager at SRI, “the vehicle can engage you in conversation to calm you down, increase the air conditioning, adjust the mood lighting, or even navigate you along routes that you tend to enjoy.” The Toyota LQ Concept Car (see video here) uses multiple, in-vehicle cameras monitoring facial expressions and body language, comparing that to a historical database onboard and using intelligent algorithms to determine anxiety versus happiness, irritation, etc. “We’re trying to create a model of preferences, habits and reactionary behavior by observing you over a long period of time. It’s a very personalized model that we build and varies according to culture, gender, age and associated behavior patterns .“ Tamrakar goes on to explain that multiple concepts have been demonstrated including integration with other senses (e.g. voice, smells, music). “With enough on-vehicle computing and driver history, all of this could be accomplished within a production vehicle and anonymized based upon data points, thereby avoiding privacy violations.”

The mentioning of a personality or conversational robot is a hot button for many: either it’s a strong want like Kristen Hall-Geisler’s “Ten Features We Want to See in Self-Driving Cars” or a visceral rejection based upon previous failures like Microsoft’s “Clippy” from 2000-2007. Having direct feedback via camera-sensed reactions from the user, though, will help to improve the system. “The in-vehicle conversation is a closed-loop system, so it will improve upon what the user enjoys. In addition, all companies have an interest in improving their system over time, so solutions we’re exploring are how to upload non-user-specific lessons learned to improve the generic model.” Tamrakar  also points out that the models can also sense nervousness in the driver’s voice, which again can help to reinforce the video-based models.

Another unrelated solution that might prove useful is “guarded introduction” of autonomous technologies. A 2018 AAA study found 53% of the public is comfortable with autonomous shuttles for airports or theme-parks, which could permit users to better understand the technology before mass adoption on unscripted highways. As evidenced by adoption and reactions from The University of Michigan students about its North Campus autonomous shuttle, such a strategy can work even on open roads if driving parameters are sufficiently benign (e.g. max speed of 12 mph). Familiarity, in this instance, would un-breed contempt.

Lastly, various companies have approached education. Waymo’s Head of User Experience Research & Design, Ryan Powell, recognizes that “… most people are experiencing self-driving technology and learning how it works for the very first time …” and that “… establishing trust with our riders … starts with communication – we need to make sure riders are well-informed.” Along those same lines, AAA has partnered with The University of Toledo to host public seminars over the past two years educating potential customers on the promising improvements of autonomy and the ongoing challenges like cybersecurity.

Strategy To Probably Reconsider

Some of the original hype around autonomous driving was about effortless, worriless commuting and passenger-enhancement features. WIRED’s 2018 article stated, “… the term ‘driverless car’ will soon seem as anachronistic as ‘horseless carriage.’” Along those lines, the #3 feature in that same Hall-Geisler list was “Naps, Snacks and Movies” with the closing line of “Get ready to text and not drive, America!” And major automotive companies like Adient, Volvo and Bentley fed into the excitement by demonstrating concepts like reclining seats that allow the driver to “lie back”, pop-up TV screens for the ultimate in relaxation and champagne-laden tables with leather couches. Even Mercedes flirted with augmented reality and providing interesting information on-windshield about various points of interest along the roadway. Ahhh … seemingly perfect comfort in a quasi-living room.

But if we have learned anything from the death of 38-year-old Walter Huang, it’s that putting the driver 100% at ease isn’t a good, near-term idea. Despite Tesla’s instructions to drivers to keep their hands on the wheel and day-over-day warnings of near accidents at the eventual barrier that killed Mr. Huang, he was imperfectly comfortable playing video games while Autopilot failed to avoid the fatal collision. And Huang wasn’t alone. Multiple reports have surfaced of drivers napping and vehicles crashing where autonomy may deserve partial blame. Now those automotive manufacturers following ISO standards and employing “state of the art” Functional Safety and Cybersecurity practices (e.g. early/often 3rd party assessments or audits) recognize that drivers must remain cautiously engaged for minimally the near-term.

So maybe a healthy amount of fear isn’t a bad thing … just as long as isn’t off-putting entirely.

LET'S TALK

Do you need to improve your automotive product development, to increase efficiency, or to comply with ASPICE and Functional Safety? You are at the right place.

Back