Categories: world

Can Tesla go out for the industry and deliver completely self-driving cars next year?

April 23, 2019 by Paul Fosse In this article I will examine the claims that Tesla made on his independence day at Tesla on April 22. Although I am a small investor in Tesla and undoubtedly a fan of the company and its cars, I try to be as objective as I can be and show where Tesla's claims are proven or uncontested and where they are ignorant and require a hope of faith. To get a robot ax, Tesla must have the following pieces of the puzzle: 1. Cars. Tesla claims it has built approximately 500,000 vehicles that have the necessary sensor input (which Tesla built since October 2016 has 8 cameras, 12 ultrasonic sensors and 1 radar), and these cars can get the upgraded FSD computer. Tesla plans to make another 500,000 cars in the next year, so it should have about a million cars available to compete with Uber and Lift on a switch. Of course, Uber and Lyft have more than a million cars registered on their platforms (thought to be over 2 million only on Uber). Not all the millions of Tesla cars are willing to let a stranger ride in them for money. On the other hand, the Uber and the Lift drivers run only a few hours a day, while Teslas can go up to 24 hours a day, because they do not need a driver. It seems clear that Tesla will not have the scale to do significant damage to Uber and…

April 23, 2019 by Paul Fosse


In this article I will examine the claims that Tesla made on his independence day at Tesla on April 22. Although I am a small investor in Tesla and undoubtedly a fan of the company and its cars, I try to be as objective as I can be and show where Tesla’s claims are proven or uncontested and where they are ignorant and require a hope of faith. To get a robot ax, Tesla must have the following pieces of the puzzle:

1. Cars. Tesla claims it has built approximately 500,000 vehicles that have the necessary sensor input (which Tesla built since October 2016 has 8 cameras, 12 ultrasonic sensors and 1 radar), and these cars can get the upgraded FSD computer. Tesla plans to make another 500,000 cars in the next year, so it should have about a million cars available to compete with Uber and Lift on a switch.

Of course, Uber and Lyft have more than a million cars registered on their platforms (thought to be over 2 million only on Uber). Not all the millions of Tesla cars are willing to let a stranger ride in them for money. On the other hand, the Uber and the Lift drivers run only a few hours a day, while Teslas can go up to 24 hours a day, because they do not need a driver.

It seems clear that Tesla will not have the scale to do significant damage to Uber and / or Lift when it first rolls out services. On the other hand, investors are trying to predict the future with their investments, and if they perceive Tesla’s history to be credible, it will do a great deal of harm to Uber and Lyft – if Tesla can scale over the next 5 years without paying a cent to the driver , it is obvious that the company will have much lower costs than Uber and Lyft, unless they can find access to millions of self-driving cars.

2nd Redundancy. Tesla needs cars that can accelerate, brake and control with electric motors. Elon and team mentioned that they have full redundancy in braking and steering (they did not mention acceleration), so they can fail in a servomotor and a brake motor and still steer and stop safely. I would think that you would like to quit and get the problem fixed, do not continue with the only steering and brake motor. Although this can be done with gas cars, most people say that the control of a car is a little easier with electric cars. This point is not really discussed by the company’s critics.

3rd Electric cars against gas or diesel cars. You can certainly build a self-propelled gas or diesel car (if you can of course find out for self-driving), but it is undisputed that electric cars have much lower fuel costs (about one quarter the cost in most areas). If you only drive a few miles a day, it tends to counter the higher initial purchase price of the electric car. If you drive the car a lot, like for 24 hours a day to maximize income, the lower cost of an electric car becomes very significant. Tesla is the only electric car manufacturer in the United States that has a large scale. It seems that car manufacturers across the industry electrify their lineups, but it’s generally questioned how fast it will happen, and even if it will happen.

Elon claimed that their model 3 engine and body can go millions of miles and that their batteries can go 300,000 to 500,000, but it is non-existent. At the presentation, Elon claimed that a new battery pack would come out next year designed for more charging cycles so that it would be millions of miles. This is ignorant, but Elon’s record with these kinds of claims is excellent. He has always delivered the promised battery performance, but not always in the promised time frame. It is believed that the maintenance costs of Tesla vehicles are much lower than gas cars and although that benefit is sometimes disputed, the evidence is quite strong that it exists.

4th Sensor array. Does Tesla have the right sensors?

No one disputes that cameras, ultrasonic sensors and radars are very useful, but almost everyone thinks that sufferers are needed. I’ve written about this. CleanTechnica has also addressed this and here. The problem is that even though lidar makes it easier to find the safe areas to drive, because it gives you a 3D map of space without using any artificial intelligence (it only lights a laser and measures the time for it to bounce) it doesn’t work In bad weather and it doesn’t help with many other problems you need to solve to do self-driving. The lasers do not help with stop signs or traffic lights or recognize bicycles or pedestrians or cars or predict future behavior of any of these three. Lidar does not help to read road signs or signs or any of the aids used worldwide to help the billions of human drivers.

Lidar is great if you just want to put a car in a science project and make it walk around on the road and not run into any stationary objects in perfect weather. Then you don’t need fancy software, you can only tell the car where the stationary objects are and count a path around them.

As you can simply see (pun intended), lidar does not help with any of the problems of complex urban environments, namely moving people, bicycles, animals, cars and trucks controlled by humans or animals that make unpredictable things in any kind of weather . For that you will need some intelligence, either human or artificial.

It is Anthony Levandowski speaking above.

5. Intelligence to understand the environment around the car. In order for you to understand the way forward, Tesla claims that you need some modest CPU and graphics processing power and a huge amount of linear algebra multiplication and power. As I wrote almost a year ago, Tesla looked at what the industry had access to fulfill its data requirements and found no one who worked on a chip that met the performance requirements (especially the ability to process a single image at a time – instead of batches of 256 images – with very low power). If you use too much force, you will significantly affect the car’s driving range.

Elon recruited a top team with experience from Digital Equipment, Intel, Apple and AMD to create a custom chip. Because they had modest requirements for processors and graphics needs, they licensed the existing designs and placed them only on their chips. However, since they had unique needs for high performance multiplication and very low power additional operations and were unable to find any acceptable solutions available for license, they designed a very simple, high performance processor. It’s a well-known truism in the semiconductor industry that you can make a chip faster for an operation if you don’t have to handle a complicated instruction set.

You see this also in the encryption of mining. If you are willing to design a chip to make the mine, you can perform the operations much faster and use less power than using the CPU or GPU to do the math operations. The reason why each currency does not have a chip that is designed to break it is that it costs a lot to design each chip and it is difficult to predict which cryptographic courses are used to repay the original chip design cost. Of course there is no problem here – if this chip solves the complete self-driving problem, there is no dispute if there is a huge demand.

I’ve heard critics claiming that Tesla is not likely to be able to design a chip that is better than the “experts” on Intel and AMD, but I find its plan viable for several reasons:

  • They employed high quality industry talents .
  • They only adapted the parts of the chip that they had unique requirements for. The licensed proven (but not leading) designs for CPU and graphics engines. This project would have been much more risky if they had adapted the entire chip.
  • They manufacture the chip on a Samsung fab. Elon can love vertical integration, but he is smart enough to realize that building a 14 nanometer lithography process that goes to a 10 nanometer process is a headache that they didn’t have to touch.

6th You need lots of training data.

There is no doubt that Tesla has many more cars that run around with cameras than any other player in the world combined. It is questioned whether they can afford the mobile data charges to send all the data back to the mother vessel. If they can’t (and it is likely that they can only send a small amount of data back), do they choose the right sample to get the cross cases they need to make the cars safe? They use the drivers to help them decide what routine video they do not need and what is special that they need to look at and train image recognition software to handle.

7th Image recognition and depth perception.

Elon made it clear in the presentation that Andrej Karpathy not only was a PhD student in artificial teaching at Stanford but he developed the very popular class learned in image recognition and is undoubtedly the world’s best expert to train neural networks to recognize images. I think some would dispute that Andrej is a top expert, but many (including me) are convinced that image recognition will be able to progress as fast as Elon claims. I have read many articles about this and it sounds likely, but it is such a big step in the ability that I cannot help but doubt whether they can make so much progress in such a short time.

I will say that an example where my skepticism was disadvantaged was Alexa’s natural language ability. I had seen 30 years of PC products claiming that they would receive speech recognition and they all took a lot of training for difficult results. Then, suddenly, Alexa (and I’ve heard Google have a good one as well) solved the problem and it seems to understand what I’m saying pretty well. It still seems pretty stupid to do complicated tasks, but it does a good job at simple.

Tesla has a good team, but this problem is just extremely difficult. This is really the area Tesla just has to prove that it works because the world will not trust them, no matter what they say.

8th Drive the car when you recognize which objects are there and where they go. This is not too difficult with the exception of the chicken fight that drivers play to try to change paths. Tesla must prove that they can find a way to be assertive enough to enter a narrow lane without causing a minor accident. This is difficult for people and it will also be difficult for computers.

Conclusion

I came from Tesla’s Independence Day, impressed by Tesla’s strategy and enthusiasm, but convinced that they will be able to pull it out next year. In my 35-year career in software development, I have seen many examples of a project that I expect to take 4 years, completed in one year with excellent leadership and programming numbers. I have also seen several projects that could have been completed in a year are interrupted after several years of unfinished, usually due to leadership who had great vision but insufficient talent to pull it off. Too complicated development processes have also killed some projects, but I do not expect it to be a problem at Tesla. Elon has developed commercial software since he was 12 – he will not let a bad process kill the project.

My opinion is that they can pull it off, but I really don’t know if they can do it next year or not. As Yogi Berra said, “Predictions are difficult, especially about the future.”

If you want to utilize my Tesla reference link to get 1,000 miles free charge on a Tesla Model S, Model X or Model 3, here is the link: https://ts.la/paul92237 (if anyone else helped you, use your code instead of mine). I encourage you to buy before the price of Full Self Driving (FSD) goes up on May 1 if you believe in Tesla’s ability to make it work soon.


Tags: Andrej Karpathy, Elon Musk, Pete Bannon, Tesla, Tesla Autonomy Day, Tesla autopilot, Tesla Full Self-Driving

About the author

Paul Fosse A software engineer for over 30 years, first develop EDI software and then develop data storage systems. Meanwhile, I have also had the chance to help start a consulting company for software and make portfolio management. In 2010 I took an interest in electric cars because gas became expensive. In 2015, I started reading CleanTechnica and was interested in solar, mainly because it was a threat to my oil and gas investments. Follow me on Twitter @ atj721 Tesla investor. Tesla reference code: https://ts.la/paul92237

Share
Published by
Faela