In addition to the love of hype, Musk's pioneering position in many fields is beyond doubt
Text | Jurica Dujmovic
Edit | Guo Liqun
Translate | Small color
At Tesla AI Day in August 2021, Musk said he would build a humanoid robot called Tesla Bot (code-named Optimus), and the prototype is expected to be released in 2022.
Tesla has made a lot of moves this year, and as the end of the year approaches, I review how some of Musk's boldest plans have progressed so far.
Tesla also released the Model S Plaid all-electric sedan this year, which can accelerate to 60 miles per hour and 200 miles per hour in two seconds, making it the fastest accelerating production car ever. Tesla's current annual production of 1 million cars, while also building a large number of factories, the company is in the limelight, the stock price is also soaring.
Let's start with Tesla's robots. The 5-foot-8-tall robot is designed to weigh 125 pounds, carry 45-pound items, and move at speeds of up to 5 miles per hour. With five fingers on each hand, the robot performs tasks that humans consider "boring, repetitive, and dangerous." To build such a robot, Tesla will need a lot of sensors, connectors, brakes, Tesla cars' self-driving technology (FSD) and some technologies in the field of artificial intelligence.
Tesla supporters have speculated that the company will take us into a "brave new world" in 2022. In 2016, when Tesla launched the photovoltaic roof Solar Roof at the filming site of the American drama "Desperate Housewives", it was also this group of people (and some media) who came out to applaud.
Fast forward to 2021, and it turns out that Solar Roof is a product that hastily launched and failed to be researched, spending a lot of Tesla money but not getting customer recognition.
Five years ago, Musk also promised to achieve "full autonomous driving" (FSD), a feature that falls far short of the most advanced level 5 autonomous driving. As of 2021, the FSD is still at Level 2, and Tesla had to disclose this in a letter sent to the California Motor Vehicle Administration.
Unfulfilled promise – self-driving taxis
Musk also made an unfulfilled promise in 2019, when he announced that 1 million Tesla self-driving taxis would hit the road by 2020. But what we're seeing is not a car that can automatically pick up and drop off passengers and make money for car owners, but a "smart call" function that does not belong to autonomous driving at all.
Soon after the "Smart Summon" feature was launched, it received various complaints from users, including inexplicable wayfinding patterns and crashes that caused thousands of dollars in damage. But Musk still insists there is room for improvement in the feature.
But this is just one example. Musk, the self-proclaimed "Technoking of Tesla," has made too many promises that have never been fulfilled, and now there are even many websites that take stock of those unfulfilled promises.
"Thunder is loud, rain is small"
The same seems to be true of Tesla robots, and Musk is still doing what he does best — overmarketing but often ends up with "thunder and little rain." He once said: "Tesla cars are semi-perceptual robots with wheels. But this is not the case, Tesla did not even do a good job of the "smart summoning" function, and did not show any perception function except FSD.
If Tesla robots are really built, such shortcomings will be more obvious. The robot's eyes will use the same camera as Tesla's Autopilot function, and a computer using FSD functions will be installed inside the torso. Therefore, this is still an FSD function that requires the driver to continue to be ready to take over, or it may lead to collisions and errors.
Finally, the operation of the vehicle is much simpler than that of a robot. To understand and interact with their surroundings, robots ( especially humanoid robots ) apply inverse kinematics, servo systems, actuators, and a host of sensors to perform more complex maneuvers.
For Tesla robots to be used in complex environments, it is obviously unrealistic for a system that is still under development or even proven to be deficient in 2022.
The expectation for Tesla robots is that they can go to the grocery store on their own, but judging by the current performance of FSDs, such robots are more likely to hit people on the road, get hit by cars, get lost while calculating paths, fall down and damage. I don't think Tesla robots pose a threat to humans, they only get in the way.
With a little less hype, Musk is still a great engineer
Getting the robot to walk in a straight line and judge the terrain is a big problem in itself, and Boston Dynamics and Agility Robotics have spent many years solving it, and while the end result is gratifying, the process is long and bumpy, which is enough to illustrate how difficult it is to solve this puzzle. Musk may be able to solve this problem, but judging by his past achievements, it is impossible to achieve next year.
Does this mean that Tesla robots will never come out one day? not necessarily. Musk often works very hard to make the projects he promotes a reality, and he does a lot better in engineering than he does in promotion. In fact, if he had less hype about Tesla robots, the tone of this article would have been completely different.
The hype about robots has pushed up Tesla's stock price, masked the concerns, uncertainty and skepticism raised by in-depth investigative news reports, reassured investors and attracted new talent to Tesla's robotics division. Musk has been a pioneer in many areas, and it can be seen that he is doing the same thing in the field of robotics. Robots have been around for many years in the consumer and industrial sectors, but Tesla's robots will be unprecedented.
In the process of Musk launching the Tesla robot prototype, he may disappoint fans because of the delay in the launch time or the failure of the prototype, but I believe that in this process he will also solve many of the current research problems in the field of robotics, and eventually make the hype of Tesla robots become a force that really promotes the development of robots.
A more realistic timeline
In the process of studying Tesla robots, Tesla may eventually become the first company to produce such robots, but the timeline proposed by Musk is very unrealistic.
So what timeline is more realistic?
In an interview with Reuters, Raj Rajkumar, a professor of electrical and computer engineering at Carnegie Mellon University, was asked the question. His response: "I can say with certainty that 10 years is not enough for a humanoid robot from any company on the planet to go to the grocery store for you." ”
Do you think Tesla has the ability to build such a robot in the short term? What are your thoughts on Tesla robots? Do humans need such robots? Feel free to leave us a message to share your views.
Copyright Notice:
Original barronschina articles, not reproduced without permission. In English, MarketWatch reported on December 9, 2021 that "Opinion: Tesla and Elon Musk have had a wild run this year. Nothing was as crazy, or as intriguing, as the Tesla Bot”。
(This article is for your informational purposes only and does not constitute the provision or reliance of investment, accounting, legal or tax advice.) ）
