#autopilot

If only Elon Musk could summon institutional questions

4381FCA6-4C35-4719-80C2-84CDEEEC1135.jpeg

Elon Musk has apparently terminated the question of a Bernstein analyst ((followed by the rest of the institutional queue) on the basis of it being “uncool”. He said, “We’re going to YouTube [for retail investors]. These questions are so dry. They’re killing me!” If only the Tesla CEO could summon the right type of questions that deflected criticism of the company as easily as maneuvering a parked Model S from a tight parking spot.

While he urged non-believers to sell the stock, there is little to be gained pushing a line of  opacity for a company with production issues, continuing losses and $10.6bn in debt. Earnings results are not about having fun but for investors/analysts to probe and qualify assumptions in the interest of making rational investment decisions.

CM has made constant reference to Musk’s amazing ability to sell. He is coming up to the pointy end of having to deliver. There are countless distractions which perculate below the surface – copyright infringement trial launched by Nikola Motor, the NTSB autopilot probe, countless resignations and recent calls to cut the staff canteen cookies. By blowing off the main investor pool that feeds him, the question of CEO capability becomes a bigger factor than the dreadful earnings themselves.

There is no better disinfectant than sunlight but Musk continues to deflect. Cash flow continues to decline  The production shutdown in April will thump Q2 earnings, not to mention the capex spend should rise plus the write off of equipment that has proven to be surplus to requirements. Here he is talking of 10,000 units a week down the line to fill the hearts of the faithful followers. Perhaps his comments about not needing to raise capital are best addressed by the fact he’s raised 7x since that statement.

Today’s results meeting is more telling in that snake oil salesman tactics of talking up the situation was replaced by silence and stonewalling. Telling.

Tesla’s Autopilot beta testing means customers are crash test dummies

DCD0B79C-8F2E-4A81-912B-A7B6EE330A61

Back in April 2016, we wrote about the dangerous legal ramifications facing Tesla due to its overzealous promotion of the auto-pilot function. What people tend to forget is the issues surrounding liability. An insurance company often covers a driver with respect to accidents – wet road, poor visibility or being hit by another driver. The insurer covers that type of damage. Yet the death of a Tesla driver in California last week was found to have had the auto pilot function on. Why should an insurer pay for damages that result from willful negligence promoted by the manufacturer itself? This is a design fault. Moreover how could Elon Musk’s legal team not suggest that he refrain from such promotion? Accidents as a result of Tesla’s auto pilot are becoming so numerous that it is hard to fathom why people put so much faith in the system, as this video highlights. They are willingly becoming crash test dummies.

DF8508D5-2491-49EB-BEA8-15C64DE165AB.jpeg

Tesla’s own website notes, “Build upon Enhanced Autopilot and order Full Self-Driving Capability on your Tesla. This doubles the number of active cameras from four to eight, enabling full self-driving in almost all circumstances, at what we believe will be a probability of safety at least twice as good as the average human driver. The system is designed to be able to conduct short and long distance trips with no action required by the person in the driver’s seat.”

The video on the autopilot webpage highlighting the autopilot function on the makes no reference to ensure drivers pay attention to the road even when the system is in use. Sounds to me like the ambulance chasers have plenty of ammunition to launch a class action. It only cost Toyota $1.2bn for the runaway accelerator issue. For a company deeply in debt with such heavy losses, rising interest rates, falling credit rating and senior departures, Tesla should be careful not to get carried away with signaling the virtues of systems that are clearly flawed.

4349AF04-8ACD-4587-A072-A4640B7E29A9.jpeg

 

Tricking the auto-pilot 73% of the time

 

So much faith is put in the hands of computers nowadays but the idea of driverless cars is still fraught with danger.  Car & Driver reports “Researchers at the University of Washington have shown they can get computer vision systems to misidentify road signs using nothing more than stickers made on a home printer. UW computer-security researcher Yoshi Kohno described an attack algorithm that uses printed images stuck on road signs. These images confuse the cameras on which most self-driving vehicles rely. In one example, explained in a document uploaded to the open-source scientific-paper site arXiv last week, small stickers attached to a standard stop sign…using an attack disguised as graffiti, researchers were able to get computer vision systems to misclassify stop signs at a 73.3 percent rate, causing them to be interpreted as Speed Limit 45 signs..”

Sure systems will improve over time but we already have a plethora of people already putting too much “blind” faith in systems being fool proof as this video demonstrates

Tesla recall – Tempting Extremely Serious Legal Action

IMG_9052.PNG

Exactly 1 year and 2 days ago I concluded that Tesla was tempting too much fate with its auto pilot function. I wrote along the lines that it’s Mario Kart type rainbow road auto pilot function was distracting so much so that owners were taking cellphone screenshots and videos while they were breaking the speed limit. Even worse Tesla CEO Elon Musk was encouraging its use and talking of its infallibility. I wondered whether his legal team had been consulted before he tweeted there were more such gimmicks on their way.

It now appears that Tesla has a 53,000 car recall on its hands and as predicted a class action lawsuit against it. While Tesla’s market cap may have exceeded the likes of Ford and GM it still doesn’t generate a profit. Car makers are the target of ambulance chasers. Toyota know a thing or two about that. How many traffic accidents in Toyota’s spiked after it was revealed they had a sticky accelerator when so few incidents occurred before.

Tesla’s autopilot has killed people who trusted it to save them. It so happens that the lawyers smell blood with a class action being launched as follows:

“The lawsuit, filed by law firm Hagens Berman on Wednesday in California’s Northern district court, said Tesla’s partial autopilot technology was advertised as safe and “stress-free,” but instead “is essentially unusable and demonstrably dangerous.”

“Unwittingly, buyers of the affected vehicles have become beta testers of half-baked software that renders Tesla vehicles dangerous if engaged,” the lawsuit says.”

Tesla cars with the Autopilot 2 features were first sold in October 2016. The first generation of the system was first unveiled in 2014. The Autopilot 2, or Enhanced Autopilot, feature costs consumers $5,000.

That is the problem with Tesla. It is a victim of its own self created bubble. It promises everything but delivers realities far removed from those dreams. Quality control is still an issue and the endless stream of subsidies has to  have an end date.

Tesla  is valued at 18x more than Mercedes Benz. Musk mentioned how he recently discovered how important production efficiency was. That is much of his battle. Even if he irons out his product glitches the incumbent makers have decades of experience in mass production, distribution and a swathe of new product lining up to challenge the space  Tesla has opened. Don’t forget Toyota sold its stake in Tesla. After  lifting the hood on its technology Toyota realized it was an empty shell. Let that sink in.

Tesla proves autonomous vehicles have a LONG way to go

IMG_9617.JPG

I’m not a believer in autonomously driven vehicles. This idea that a computer, if pre programmed, can outsmart a human. Sure, the raft of new safety features (auto brake assist) and lane assist control etc can help in situations when people cruising at brain dead slow speeds are busy texting and checking FB. Yet, there is a point where these systems are dangerous. I have driven cars with them and there have been times where the car outputs are the exact opposite of my inputs. It is unsettling and downright dangerous so I tend to switch these aids off. This excerpt from the Tesla Owners forum on FB shows how the latest and greatest auto-pilot function is flummoxed by such a simple situation. Read on.

Found a bug in 8.1 the hard way. Ruined two rims after 15 minutes of use.
That’s what happened yesterday: I started the AP on a smaller street with a sidewalk with a curb on the right. There was no line on the street next to the curb, but a line for bicycles on the sidewalk. The AP then suddenly pulled right, as it was irritated by the line on the sidewalk and ignored the curb. The rims touched the curb before I was able to react, even though I had my hands at the steering wheel…I already posted this in a German group yesterday and some people told me they had the same situation, but were able to react before it was too late.”

The idea that people put complete faith in auto-pilot systems is a worry. By the same token more advanced systems are supposed to use inbuilt algorithms to determine whether to swerve away from the kid on a BMX bike doing skids on the sidewalks toward the edge of the kerb braking as late as he dares and an old lady on a crossing 5 meters further on. The system may choose to sacrifice you the driver, err sorry passenger. While there is no doubt autonomous systems will continue to get better, would you prefer your airline pilot to be limited to a computer software program only or would you prefer a human in the cockpit who can assess the situation in real time?

Maybe I’m too analog. A fuddy-duddy that refuses to accept the future. I don’t think I’m alone but one day more people will grow tired of an app-driven existence. Life will become too boring and they’ll soul search for more tactile experiences. I was tinkering in the garage on my bikes fitting new parts, tyres, cleaning chains and doing oil changes. There is a something to be said about zen and the art of motorcycle maintenance. I was completely at peace after completing these analog tasks because it requires a focus that can’t be found in a 15 second swipe of an app.