Playing The Elon-g Game: Where Tesla May Have It Right And Wrong

Friday, April 24, 2020

#Functional Safety    #Automotive SPICE    #Automotive Agile

It has been a tumultuous few months for Tesla TSLA and Elon Musk. Within this bumpy ride, there are some positive and negative Lessons Learned for both engineering and business decisions. Companies wanting to compete and differentiate in autonomous, electric and innovative vehicles need to pay attention.

But first, a quick reminder of where we’ve been:

Nov 22, 2019: Tesla unveils the revolutionary Cybertruck, and Elon has to explain away the accidental smashing of the Armour Glass by a steel ball. Stock drops from $354/share to $333.

Nov 23, 2019: Musk announces 150,000 preorders of the Cybertruck despite embarrassment in the press.

Jan 29, 2020: Tesla announces just a 2.2% year-over-year revenue growth and 25% fall in net income, but the stock soars by more than 40%.

Feb 11, 2020: Tesla’s stock shoots above $800 after Morgan Stanley     MS said the “…tectonic shift to electric vehicles … could be valued at $25 trillion over the next few decades.” 

Feb 13, 2020: Tesla faces a new subpoena from the Securities & Exchange Commission (SEC), and announces that it plans to issue $2B in new common stock since they are “…actually spending money as quickly as we can spend it sensibly.”

Feb 25, 2020: National Transportation Safety Board (NTSB) finds Telsa’s Autopilot feature partially culpable for a fatal crash. “If Tesla doesn’t add driver monitoring safeguards, misuse of Autopilot is expected and the risk for future crashes will remain,” the board wrote in one of its findings.

Feb 20, 2020 to March 20, 2020: The Grand Princess cruise ship pulls into San Francisco’s bay, the COVID-19 virus spreads around the world, California’s Governor (Newsom) announces a shelter in place mandate, and Telsa’s stock drops 52% during that span versus the S&P dropping 32% during the same month.

April 2, 2020: Tesla reports better-than-expected 1st quarter sales, and the stock temporarily jumps above $500/share.

Okay, so now that we’ve taken that trip down Memory Lane, what should those Fast Followers looking to jump on the connected, autonomous vehicle learn? Three major lessons:

1. Being Agile Means Good Business

Although the amorphous definition of Agility gets even more confused among those simply lobbing buzzwords, the masses have generally understood and equated agility with “fail forward fast”, quicker times to market, better User Experiences (UX) and/or Customer Experiences (CX), etc.  Silicon Valley has certainly shown the world how to apply the Agile Manifesto towards websites and applications, and has flirted with using it on safety-related systems such as planes, trains and automobiles.

The lesson that Tesla is showing is one that was clarified in a 2014 Forbes’ article: The Only Innovation Metric That Matters: Speed to Truth. “How can you more quickly and cheaply reveal the truth across all of your assumptions, and how can you prioritize these assumptions to validate the ones most critical to the project as soon as possible? Speed to truth is the true measure of progress in the discovery phase of a new growth business.”

So Agile buzzwords aside, what Tesla has successfully done is gather rapid feedback on business and engineering assumptions.

2. Ignoring Functional Safety is Bad Business

“Foreseeable use of the product” and negligence have been well-known terms to manufacturers in all industries and global courts over the past century or two, but certainly the automotive industry has seen its fair share of defective designs and resulting suits. In 1842, Winterbottom vs. Wright was about a defective postmaster-general’s wheel that injured Winterbottom. More recently, the 1956 Volkswagen’s rear axle and suspension allegedly caused the car to roll under certain conditions and the jury found for the plaintiff causing a recall on over three million vehicles. Or even in the last few years, large cases and/or product liability recalls include the Firestone Tires, the General Motor’s ignition switches, and the Takata airbags. All had significant long-term affects to the company’s profitability and brand.

Maybe the largest lesson from Tesla’s Safety Of The Intended Functionality (SOTIF) incidents has been the foreseeing misuse of the product in combination with marketing of an “AutoPilot” functionality. Drivers find ways to nap behind the wheel of Autopilot thinking that they need not act as a driver despite warnings. And none of this is new to the industry: Urban Legend has it that “Cruise Control” has been mistaken repeatedly since the 1970s as a substitute for AutoPilot, including a possible claim paid by AllState    ALL  in 1987 for a flipped van. Regardless of the true history, Ford changed the name to Speed Control in the 1980’s to avoid user confusion.

And so, here seems to be the conflict: SOTIF together with Functional Safety requires discipline, and both are complicated standards requiring a set of serial processes to be implemented and documented. Post-It Notes® and white boards seem to be the antithesis of the regiment required. At face value, it appears as though Agility and SOTIF/Functional Safety are antonyms; go quickly versus go methodically.

3. The Fulcrum is a Delicate Balance

Gödel's incompleteness theorems suggest that every non-trivial (a.k.a. “interesting”) system is either incomplete or inconsistent. Thus, there will always be questions that cannot be answered. If the goal is to be 100%, it will never be achieved. Maybe the authors of Automotive SPICE® got it right in this regard: You can get a “Fully Meets” without achieving 100% capability.

Therein, SOTIF, Functional Safety and Automotive SPICE® practices can be enabled smartly within the designs of Agile Development as long as they are prioritized. As said well by Claudio Gregorio, “[Agile Development’s] ability to create and respond rapidly to change may appear to be diametrically opposed to the requirements of functional safety, but [the two worlds] do not have to be mutually exclusive.” The best Agile practices are about prioritizing between important tasks, communicating about transparent work, and adapting to the marketplaces’ needs. It’s not about throwing away processes, but rather about breaking them into consumable parts and concentrating on the urgent ones.

It’s also worth noting that overhyped, poorly-understood “fail fast, fail often” is supposed to be small failures with quick adaptation. As correctly noted by Dan Pontefract, “’Fail fast, fail often,’ is not only being used incorrectly as a cousin to ‘Lean’ and ‘Agile,’ it is creating a culture of people aiming for the short-term, living in a world of frenetic bedlam. Instead of calmly and intelligently iterating, employees race to complete something (failing) while racing to the next objective as quickly as possible (failing, but quicker).”

Maybe Tesla’s recent history is showing us that playing the Elon-g Game should be measured by “Speed to Safe Truth”. 

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