[ad_1]
Consumer Reports announced Thursday that Cadillac Super Cruise was leading its tests of semi-autonomous driving systems. Tesla's autopilot was second, with Nissan's ProPilot driver coming in third and Volvo's Pilot Assist coming in last.
CR's results matched mine, but certainly, even though I compared Super Cruise and autopilot, their tests were not as exhaustive and technical as those undertaken by CR.
There is however a bigger problem at play here. Super Cruise and Autopilot are both advanced extensions of the traditional cruise control system, which has existed since the 1960s and effectively locks the vehicle's throttle. The driver must be vigilant and regain control of his vehicle if necessary. In practice, the cruise control is only a highway, which allows the driver to rest his foot a moment.
Sensors, radar, laser-radar mapping, GPS and computing power have created a new world where cruise control can be greatly enhanced. Adaptive cruise control has allowed to accelerate and slow down depending on traffic conditions. Now, Super Cruise and Autopilot allow hands-free operation, though, in my opinion, only Super Cruise succeeds. (To be fair, Super Cruise does not deliver in slow traffic, a place where the autopilot shines.)
Iterative or radically new technologies
But in the end, all these systems are iterative – they build on what used to be the case. And in fact, Super Cruise is a semi-autonomous driving technology for the consumer that can be installed in today's vehicles. operating in very specific circumstances.
The autopilot is a different story. Tesla thinks that it can be developed to work as a totally autonomous system. At some point in the future, you can ask your Tesla, from afar, for example your garage in Palo Alto, to go to the San Francisco airport to pick you up. This would then take you home when you check out Elon Musk's Twitter feed without ever putting a hand on the steering wheel or giving your feet the trouble of pressing a pedal.
Metaphorically, this means that Tesla thinks the autopilot is a step-by-step function. And although Tesla cars can not achieve level 4 autonomy (autonomous driving with some parameters, but with the traditional controls still in place), by integrating autonomous hardware to each new vehicle, the company thinks it can solve the problem later, once the software, machine learning, network effects, etc. caught.
That's good, but other companies that want to achieve full autonomy take a completely different approach.
There is no way to erase it. As impressive as autopilot either – and for a builder like Tesla who has been making vehicles seriously for about five years, it's really impressive – Tesla is far behind the General Motors and Waymo's cruising effort of the Alphabet.
GM Cruise and Waymo do something completely different
I want to stress right away that this is not a problem for Tesla. The most enthusiastic supporters of society seem to think that it will change paradigm on all fronts, which is good. But solving the problem of level 5 autonomy – no human interaction with the vehicle – is an expectation on the part of a company that has about $ 2 billion and is in the process of model 3, which will be the bestseller of all – electric vehicle of all time, if the current sales trends are maintained and Tesla can solve its production and delivery difficulties.
The truth is that Elon Musk is a fantastic salesman, with a real understanding of what the customer wants. A decade ago, it was sexy electric cars. Nowadays, it is the cars that can drive themselves. To maintain the incredible mindset of Tesla, which represents billions of dollars a year in advertising on which Tesla spends nothing, Musk must keep society in the foreground with the coolest material and the most more high-tech. It does not matter that the manufacture of cars on an industrial scale has nothing to do with the manufacture of autonomous cars, even on an experimental scale.
There is money to do both. But we know the limits of standard cars: an exceptional performance is 10%. Many major manufacturers are struggling to achieve this. Luxury brands can do better, but the luxury market is unstable.
Tesla's goal is to produce all-electric vehicles with a much larger margin. This has led all kinds of observers to conclude that a complex machinery manufacturing business weighing more than a ton would reflect Apple's epic success and 30% margin. , a company whose basic product costs a few hundred dollars to manufacture. and slip into your pocket.
Established automakers understand the limitations of their operations, whose financial rules have not changed much in the past 100 years, and are eager to turn to lucrative, high-margin service delivery while utilizing the flow of money. Huge cash but limited by profits from legacy operation finance the transition. Tesla has also made gestures in this direction, but the "Tesla Network", sometimes discussed, remains the vaporware.
Morgan Stanley recently valued Waymo at a speculative amount of $ 175 billion, while SoftBank's and Honda's investments in GM's Cruise brought the value of this division to $ 14.6 billion, representing roughly more than a fifth of the market capitalization of GM $ 50 billion. So, Tesla is not even in the catch-up phase; In fact, it was dubbed two or three times by GM and Waymo even before leaving the starting blocks.
DO NOT PANIC
Before fearing that GM and Waymo will crush Tesla and cancel the autopilot, rest assured Tesla's ability to create, as soon as possible, the three best 100% electric vehicles ever built while simultaneously validating a major market for vehicles. electric that did not exist. a decade ago.
One of the regular intellectual mistakes made in Silicon Valley companies is to assume that because they have been remarkably exceptional in creating software value, they will equally succeed in creating tremendous value in. .. everything else . So, is the Tesla scam the new Apple, or even the assumption that Apple is somehow preparing to do something spectacular in the transportation sector while the company is just floundering with his project (hopefully) Apple's secret car.
The result here is that Tesla would be wise to continue to improve the autopilot – because the autopilot is great, deserves to be improved and will help Tesla to sell more cars. But Tesla would also be wise to no longer think of the autopilot as a stepping stone for something that the autopilot is not and could never be.
We already know that it costs GM about $ 5 billion for Cruise to move into the prototype stage to offer a fully autonomous car-sharing service for the big cities, which can be carefully mapped. . Recently, GM has sought out large external investors to meet the costs and risks.
If Tesla wants to pick up speed with Autopilot, it would be wise to do the same – and since Autopilot is not a separate entity from Tesla, it could use Autopilot's market share to open a new channel financing. If you think about it, Tesla would be even wiser to create a Tesla Mobility division to attract outside investment, which would allow Tesla to raise funds in a way that is not as controversial for critics who think Musk is a specialist in the destruction of capital.
"It's extremely difficult to come up with a general self-driving solution that works well everywhere," admitted Musk last week when Tesla announced an autopilot update that will be lacking, at least for the moment, full integration between technology and Tesla's navigation systems.
He is right, but what was not said is that it is even more difficult to apply the solutions already provided by the autopilot to problems that go beyond that. Tesla's more pressing problems – mainly the construction of new model 3 vehicles – outweigh everything at stake, which is far from solving the problem of the futuristic fantasy of autonomous vehicles.
Source link
