Friends who play Texas hold’em should be familiar with the term “implied odds”, which means the size of the implied future success probability. There are many articles analyzing various data of Tesla in the market. Here, we will talk about those “data” that are not so well-known, and these investments are the key to Tesla’s long-term success.
Elon Musk, the founder of Tesla, believes in starting from first principles to think about all problems, even taking a unique approach to his children’s education, which focuses on a “problem solving mindset” in the school he founded. If even the next generation is educated like this, I believe that Elon must have deep faith in this methodology. This echoes Steve Jobs’ advocacy of rethinking common things and focusing on why, not what.
What advantages does the “everything as a product” first principle bring to Tesla?
Next, let’s take a look at several points that Tesla has thought about from first principles that are critical to whether the vehicle can sell well in the long term.
Tesla’s investment in hardware development is very large, which refers to the self-developed components. Its logic is to self-develop all core components and components that are highly related to future core values, or move towards self-development based on existing solutions from suppliers. I believe that Tesla’s self-developed components are mainly concentrated in several core value points that are critical to whether the car can sell well in the long term, mainly involving automatic assisted driving modules, modules related to cruising range, automobile production and manufacturing methods, and batteries.
Self-developed Automatic Assisted Driving Module
Firstly, there is the automatic assisted driving module, which is a core function that can bring a whole new world through long-term evolution. The future culmination of this will achieve fully automatic driving, be able to ban drivers, greatly increase the time that vehicles are on the road, reduce idle vehicles in parking lots, and thus greatly reduce the overall social cost of travel. Even if fully automatic driving cannot be achieved in the short to medium term and only automatic assisted driving can be accomplished, and in some situations, drivers need to take over, it can also free up the energy of most drivers, and also reduce accident rates in daily driving, which is a very strong ability in the process of daily use. This is the core part that contributes to user stickiness.
Modules related to automatic assisted driving, such as sensors and chips, Tesla has replaced components with large improvement space from outsourcing to self-development, such as chips. Starting from early 2019, all of Tesla’s vehicles have been equipped with HW 3.0, in which all chips have been replaced with Tesla’s self-developed FSD chips, with overall computing power increasing by more than 20 times compared to the previous generation. It is rumored that HW 4.0 has completed design and began to be produced by TSMC; it is expected to be installed in vehicles in Q4 2021, with its computing power more than three times that of HW 3.0. Mature sensors such as cameras, millimeter wave and ultrasonic are outsourced.The noteworthy point is that Tesla’s long-term autopilot technology development is based on the visual route, utilizing human-like evolving patterns to enhance its ability. The vehicle’s sensor layout has a high requirement in this scheme. The core of machine learning lies in the accumulation of data and further training of deep learning networks to achieve high reliability statistically. Therefore, the layout of sensors and even the differences in pixels may have a certain impact on the coherence of data and the universality of trained networks. Therefore, according to my understanding, the reason why Tesla keeps its sensor scheme stable is that it is based on a long-term exploration to find an optimal solution and is also a responsible scheme for maintaining the coherence of the training network system. In short, Tesla has areas of rapid change as well as areas that remain unchanged for a long time, mainly depending on the underlying profound meaning.
In-house Development of Range-related Modules
Secondly, modules related to range have a direct impact on electric vehicle sales. Generally, users first look at whether the range can meet their needs, and then consider other functions. The range-related modules include the motor, power management, and thermal management.
Tesla has always self-developed the motor system, but the efficiency of motors on the market is already above 90%, and there is limited room for improvement.
Tesla’s power management system is also self-developed, including the management of self-developed battery charging and discharging, which is also crucial for the long-term battery life. If the full discharge and full charge of the entire battery block during regular use causes damage to the entire battery block, and the entire battery block is composed of individual battery cells, then the full discharge and full charge of a single battery cell is damage to individual batteries. Therefore, it is necessary to manage the status of all batteries, and only by balancing the use of each battery cell as much as possible can the attenuation rate of each battery cell be delayed as much as possible.
One of the keys to Tesla’s thermal management system is the self-developed heat pump because the charging and discharging of electric vehicle batteries have a working temperature limit. Heating is required for stable battery operation in low-temperature situations. The working principle of the heat pump is to collect the heat energy from the charging and discharging of the battery, the heat energy in the air environment, and the heat generated by the motor operation, and then distribute it to the battery heating and the cabin heating. Therefore, the management of all heat sources in the entire vehicle is one of the keys to increasing range. Data shows that the heat pump can save about 10% of electricity, which is an increase in range of 10%.
In-house Production and Manufacturing Mode
Thirdly, Tesla’s production and manufacturing mode. Traditional car production is a very traditional industry, and Toyota is the best car company among those that produce millions of vehicles, with a vehicle production gross margin of about 18%. Tesla has proposed the concept of “Gigafactory as a product”, which means that the entire factory is an innovative product that can break the traditional manufacturing mode to create higher production efficiency. Because Tesla’s factory has little external information, I guess that Tesla is creating higher production efficiency through several factors.The first step is to reduce parts and assembly procedures in order to simplify product design and vehicle design by merging modules, simplifying wiring harnesses, and simplifying interior trim. At the same time, simplify production and assembly procedures, for example, the Model Y’s integrated die-casting machine reduces the number of body parts from 70 to 2. In addition, increase the commonality of parts between different product lines, further reducing the variety and types of parts, and increasing purchasing bargaining power. For example, although the Model 3 and Model Y are different product lines, their parts commonality rate is above 70%. This may also allow the sharing of some production lines, facilitating the adjustment of product line capacity according to market demand. These innovations make vehicles more adaptable to large-scale production, while also quickly reducing manufacturing costs.
The fourth aspect is the battery. Tesla unveiled its latest self-developed 4680 battery design on Battery Day in 2020. The battery has been optimized in terms of cell size, internal structure, and heat dissipation layout, greatly improving the thermal efficiency of battery charging and discharging, while also reducing the number of cells and accessories and lowering the difficulty of cell management. The greatly improved thermal efficiency of the battery has a decisive role in increasing the supercharge rate and durability of the battery at high speeds.
These four aspects have laid the foundation for Tesla’s hardware strength. The self-developed components, including the autopilot module, the range-related module, the automobile production manufacturing method, and the battery, ultimately contribute to improving the user’s driving experience and lowering the vehicle’s selling price.
Self-development of a single component may bring certain advantages to a company, but Tesla has adopted a self-development approach for all critical components, which can create significant advantages when gradually incorporated into products. This is also one of Tesla’s key implicit probabilities – Everything as a “complete” product, which may pay back several times more than optimizing individual components from a systemic perspective.
This reminds me of the unique skill of the Seven Masters of the Quanzhen Sect from Jin Yong’s martial arts novels – the formation of the Big Dipper Array.
This article is a translation by ChatGPT of a Chinese report from 42HOW. If you have any questions about it, please email bd@42how.com.