On June 22, 2025 (local time), Tesla launched a robotaxi business in Austin, Texas, USA, using driverless automated driverless vehicles (ADVs).
The robo-taxi service will utilize approximately 10 current Model Y vehicles, and the company plans to conduct demonstration tests and accumulate data until 2026, when Cybercab, a dedicated robo-taxi model, will be officially launched.
Tesla’s robo-taxi service will be offered at a flat rate of $4.20 (approximately 619 yen, US$1 = 147.4 yen), is not available to people under 18 years old, and is available in certain areas of Austin.

Tesla’s launch of the robo-taxi business has several important implications, the most notable of which relates to the vehicle system.
Companies such as Waymo and Zoox, which have been demonstrating robo-taxis for some time, have utilized sensor fusion to achieve Level 4 SAE standards or higher for automated driving.

These companies’ ADVs are equipped with a variety of sensors, including cameras, radar, LiDAR, and ultrasonic sensors, and use this information to perform 360-degree sensing around the vehicle to achieve automated driving.
However, this method can lead to very high vehicle prices: ADVs are based on BEVs (electric vehicles), which are considered to have advantages in terms of control and system structure, but the price per vehicle increases even more when LiDAR and other expensive sensors are installed.

Tesla, on the other hand, uses 360-degree sensing with eight cameras (similar to the human eye), and the system makes decisions based on visual information.
In addition, the driving data collected from approximately 7 million Tesla vehicles deployed around the world and the supercomputer “Dojo,” which learns this information, are linked to the onboard AI chip in Tesla vehicles to realize automated driving.
In terms of vehicle cost, Tesla has already achieved mass production and in-house production of sensors and software for ADV (and AI technology), which puts it in a better position than other OEMs and robo-taxi service providers to provide relatively inexpensive services.

Another notable aspect is software.
Tesla’s self-developed Tesla OS and Full Self Driving (FSD) software are linked to Tesla’s existing OTA (over-the-air update) and connected platform, and are said to enable more advanced and precise driving control through software updates. The Tesla FSD is said to enable more advanced and precise driving control through software updates.
In fact, many users who have experienced the Tesla FSD have commented that it “drives like a human,” and Elon Musk has indicated that he aims to achieve human-like self-driving by having the AI learn human behavior.
With this approach, it is believed that software that evolves through repeated AI learning will further demonstrate its true value through OTA updates.

As the Tesla case shows, ADV is splitting into two directions: traditional hardware-centric technology and software-centric technology.
While many OEMs are currently focusing on hardware-based ADV development, there is a clear movement toward software-based ADV development, especially among emerging BEV manufacturers and Chinese OEMs that are following Tesla’s lead.
In any case, there is no doubt that we are on the verge of the realization of ADVs that transcend the traditional concept of personal ownership and specialize in sharing and service, and the coexistence of humans and robo-taxis.