news

(this article is mainly translated from gail alfar's article on x)

1.

everyone is excited about the arrival of the robotaxi. in the parking lot of warner bros. studios in burbank, california, netizens have already witnessed a vehicle wrapped in yellow camouflage and without rearview mirrors driving autonomously.

this car is likely to be used for the october 10teslapeople at robotaxi unveiling events offered demos and even rides.

what we currently know about tesla robotaxi:

a. it will be mass-produced at the texas gigafactory (elon mentioned this)

b. at the texas gigafactory, it will be strongly supported by a large-scale computing cluster called "cortex".

on august 6, elon shared an internal video of cortex, a brand new ai training super cluster located in the texas gigafactory. the video shows nvidia h100 and h200 gpus. cortex will have about 100,000 such gpus, which will provide powerful power for the al training required for tesla robotaxi.

elon: today’s video inside cortex – the massive new ai training supercluster being built at tesla hq in austin to solve real-world ai problems

based on several tesla patents, we can speculate on some of the characteristics of robotaxi:

a. the robotaxi will be supercharged wirelessly.

b. the robotaxi will disinfect itself.

c. robotaxi is ready for fleet management

2. wireless charging

tesla has a wireless charging patent that can be viewed on the wipo website. according to notateslaapp:

the first of the four patents covers temperature sensors and wireless charging applications. the second patent is a shorting switch designed to prevent or reduce energy leakage to ground sources near the wireless charging mat. the third patent is actually the wireless charging circuit itself, and how to manufacture it. the last patent is for intelligent estimation and adaptation to changes in wireless charging layouts and systems.

3. automatic disinfection

we can also speculate on the cleaning system of the robotaxi through a patent recently obtained by tesla. in june, patent researcher seti park announced the details of this amazing system at x.

tesla's patent wo2023/163943a1 proposes a system and method for automatically disinfecting enclosed spaces that can be shared by multiple people, particularly the interior of a vehicle.

the core of this patent is to detect the current environmental conditions in the enclosed space, generate and execute a cleaning process based on these conditions. this cleaning process is designed to utilize various components within the vehicle.

key features of the system include:

a. environmental monitoring through various sensors: detecting temperature, humidity, presence of pathogens, etc.

b. personalized cleaning process: generates appropriate cleaning process based on detected environmental conditions.

c. various cleaning methods: utilizing hvac air conditioning systems, ultraviolet lighting, steam generators, etc.

d. automated process: automatically perform cleaning process when people are not around.

e. adaptive system: adjusts cleaning methods based on external weather, vehicle usage history, etc.

4. fleet management

waymo, an autonomous driving company that has been operating in many places in the united states, cooperates with uber to manage its fleet, but tesla’s recent patent wo2023/069635a9 shows that they may not need such cooperation.

the patent describes an advanced ai-driven system capable of managing thousands of vehicles simultaneously using anonymous data collection and predictive maintenance. the privacy-preserving technology appears to be tailor-made for large-scale robotaxi operations, suggesting that tesla is well prepared to launch a standalone robotaxi service.

the core innovation of the patent:

1️⃣anonymous vehicle data collection and processing

what it does: collect and analyze vehicle data using anonymous identifiers.

benefit: enables large-scale fleet monitoring while protecting individual vehicle privacy.

2️⃣ai-driven vehicle warning technology

what it does: predict potential problems based on data collected from multiple vehicles.

benefits: preemptive repairs improve overall fleet reliability.

3️⃣ targeted instruction distribution

function: send instructions to specific vehicles based on the warning results.

benefits: effectively manage large fleets without compromising privacy.

4️⃣based onmachine learningprediction

purpose: continuously improve forecast accuracy based on collected data and results.

benefit: increases the system's ability to identify and resolve potential issues over time.

5. 

gail alfar says:

i use tesla fsd 12.5 daily on complex construction sites and busy areas within austin, texas. the tesla vision system drives better than most, if not all (in fact, it’s hard to say for sure that all).

tesla's vision system only knows one thing: to get from one place to another safely. yet, people use their brains all the time to do all sorts of things. there's nothing wrong with that; it's human nature and a fact of life.

one day, we will look at robotaxi the same way we look at elevators today. we get in, press the button for our floor, and then walk out. robotaxi will be the same.

image: teslaconomics