Tesla is leading with its insights with its software:
1. Auto upgrade is under emphasized.
2. Self driving hardware is already working hard, without doing the driving.
These two combined is advancing Tesla's software in ways others simply cannot.
Curious: Is Tesla working on a "remote driving" feature for Tesla? I would be a user for sure.
Thesis & Catalyst For Tesla Motors (NASDAQ:TSLA)
When I first started to working as a programmer, I wrote software for embedded devices. My boss was a veteran electrical engineer. Interestingly, whenever there were feature change requests, he would push me to come up with a software solution. Even if the solution I cam up is cumbersome or even backward in my opinion, and even if I believe a hardware solution will be much easier and simpler, he will make me to change the software. He would then happily exclaim to our customers that the problem well handled and "it is just software".
Being a well educated software developer, I argued with him on the pitfalls of his statement, but to no avail. Only later, after I practiced Agile software engineering, did I appreciate the differences in between hardware and software and the merits of his statement.
I believe these understandings are the advantage Tesla has over the other auto makers. Unfortunately, I did not see any evidence that the other auto makers understand them. Because of this, Tesla's lead will only going to expand.
The differences between software and hardware are two folds.
1. The variable costs for an improved software is zero: Software can be copied, period.
2. Software will never be completely "done" in the sense that there can always be improvements. In fact, many software developers would even argue that, with a few exceptions, it is futile to design the complete software and that the way to development good software is to lay a foundation so that the software can be continuously improved upon, especially when the requirements of the software is not well defined. Clearly, the software for the self-driving car is in this category.
I have two reasons to believe Tesla understand these two points well:
1. From the very beginning, Tesla cars come with the capability of automated software upgrades. Occasionally, one reads about the wonders Tesla was able to achieve through this feature, but its importance to Tesla and to anybody working on self driving cars has not been emphasized enough.
Mind you, get this right is not an easy task, especially if it was not designed from ground up. Case in point: So far, in spite of its best efforts, Microsoft still hasn't got its auto-upgrade completely right.
It is a surprise that we haven't heard other auto makers following suit. Don't they realize that this feature has allowed Tesla to deliver self driving cars ALREADY? Pending a switch to be turned on, literally!
2. Tesla has built into their cars all the hardware needed for a self driving car. Analysts, especially the bearish ones, have ridiculed Tesla for doing this before the software is available and claimed the owners of Tesla are suckers to pay for them. That is unfortunate because they failed to realized that half of the self driving hardware are sensors and data collecting mechanisms. With these sensors. Tesla is collecting data which their engineers can use to figure out the requirements for their software. Without this body of data, nobody can possibly "design" the software, period.
I also have to point out these advantages Tesla have are also feeding upon themselves: The better they get with their software, the better their data. The better their data, the better/faster they can improve their software, while others are still "designing" their self driving cars.
As an example to highlight the differences between software and hardware, I was actually curious if Tesla is working on a "remote driving" features for their cars. I can see many occasions where I have to drive somebody to places where I myself is not needed. Wouldn't that be a cool virtual reality project? That probably can be done much sooner than the self driving software and create another venue of data gathering.
Disclosure: I am/we are long TSLA.