In the past few years, autonomous driving technology has moved from science fiction movies to reality, especially the commercial application of unmanned taxis (Robotaxi), which is quietly changing people's imagination of transportation. From Waymo allowing ordinary people to call a taxi in Phoenix to Tesla launching Robotaxi testing in Austin, vehicles without drivers have appeared on the streets of the United States.
This is a technological revolution, and also a comprehensive test of systems, ethics, and business models. This article will help you sort out the current status of global autonomous driving development, analyze the technical routes, major players, and key challenges, and explore how it can move from experimentation to implementation. Let’s read on!
3 key things to take away if you only have one minute
- Autonomous taxis are now on the road, but they still have a long way to go before they become universally available.
Waymo, Cruise and other companies have launched Robotaxi services in several cities in the United States, allowing ordinary people to get on the bus without a driver. However, most of these services are still limited to certain areas and speeds, and human intervention is still required in special situations. In other words, there are technical advances, but it will take several years and institutional coordination to achieve "can be called anywhere, on call, and arrive at any time." - The technical routes chosen by different manufacturers are very different, and it is difficult to say who is right and who is wrong.
Some companies, like Tesla, advocate relying solely on cameras and AI for visual recognition, while others, like Waymo, have invested heavily in building radar, LiDAR and other sensor arrays. Both sides have their supporters and challenges, and it's like a battle between "all about the brain" and "full sensory upgrades," and there's no clear winner yet. - The real focus of competition is who can master the data and decisions behind the self-driving car.
Robotaxi is not just a car running on the road, but also a complete system of data collection, judgment and decision-making. From passenger preferences to road information, from accident responsibility to privacy disputes, the governance logic behind these will determine which player can lead in the long run. Developers and policymakers must think clearly: How much power are we willing to give to an AI car?
What is a Robotaxi and why is it in the spotlight now?
Robotaxi, as the name suggests, is a "robot-driven taxi". It combines autonomous driving technology, a shared transportation platform and a real-time navigation system to replace human driving and provide all-weather, low-cost mobility services. At first glance, it sounds like Uber, but without people in the car.
But why has Robotaxi become the focus of the industry and the media now? There are several reasons:
- AI models (especially perception and judgment systems) have made rapid progress in the past three years, especially in image recognition and real-time traffic condition processing.
- Secondly, the changes in remote work and urban mobility patterns after the epidemic have increased people’s demand for “arriving safely without driving”.
- The investment circle's anxiety and expectations about the autonomous driving business model have also prompted more companies to put their products on the streets for testing.
According to reports from Emerging Tech Brew, WSJ, Reuters, etc., the most active commercial test sites for Robotaxi are currently in the United States, including Silicon Valley, Phoenix, Austin, etc. Among them, Waymo has opened 24-hour service and has completed hundreds of thousands of driverless pick-ups; although Cruise encountered regulatory challenges due to an accident in San Francisco last year, it is still actively preparing to restart testing. Although Tesla has not yet entered fully automatic Robotaxi operations, its technology strategy of relying only on cameras and neural networks (Vision-only) has successfully attracted the attention of the market and investors, and announced that it will officially launch Robotaxi services at the end of 2025.
In addition, startups including Zoox, Motional, and Aurora are also conducting different forms of commercial operations and closed tests. Overall, although Robotaxi has not yet been fully popularized, it has entered an irreversible stage from the implementation of technology, policy coordination to changes in people's usage habits. In the next few years, the performance of these pioneers will determine whether Robotaxi can truly enter the mainstream transportation network.
Strategic differences among manufacturers regarding self-driving technology
When we talk about "autonomous driving", we often think that all car manufacturers are taking the same path.
But in fact, major global companies have adopted completely different strategies and technical architectures on how to achieve safe and commercially viable autonomous driving.
- Tesla: Visual Minimalism
Tesla is one of the few manufacturers in the world that advocates self-driving cars that rely solely on cameras and neural networks. Elon Musk calls this approach "Vision-only autonomy," meaning that LiDAR or radar is not used at all, and only cameras are used to capture images, and AI models are used to instantly judge environmental conditions and make decisions. Tesla believes that this approach is closer to human driving experience and is more likely to be expanded to diverse scenarios around the world.
The advantage of this approach is low hardware cost and flexible deployment, but the disadvantage is that it is highly dependent on model accuracy and data volume, and is prone to failure at night and in bad weather. Tesla optimizes its Full Self-Driving module (FSD) by continuously collecting driving data from users.
- Waymo, Cruise: Sensor-centric conservatives
In contrast to Tesla, Waymo (a subsidiary of Alphabet) and Cruise advocate that "multi-sensor fusion" is a necessary condition for safe self-driving. This strategy involves using a combination of LiDAR, radar and cameras to allow the vehicle to accurately detect surrounding objects even when the field of vision is limited.
Waymo places particular emphasis on "redundant perception systems" and accurate maps, while Cruise emphasizes the dynamic prediction and real-time adjustment capabilities of urban streets. This path is technically complex and costly, but it is currently more convincing in terms of commercial implementation and regulatory review.
- New startups create hybrid strategies and modular competition
In addition to the giants, startups including Zoox, Aurora, and Motional are trying to build modular, integrated system architectures. They may work with car manufacturers to provide an "autonomy stack" that other OEMs can quickly integrate into their own models.
This type of strategy is closer to "operating system platformization", intending to become the Android or Windows of the future Robotaxi industry. Whoever can build a self-driving platform that is highly scalable and easy to maintain will be able to control the core nodes of the ecosystem.
Different technical routes reflect different bets by companies on risks, costs, and timelines for scaling up. Although there is no clear winner at this stage, this three-way battle of "perception vs. judgment vs. integration" is the most critical battlefield for Robotaxi in the next five years.
Tesla's strategy and controversy: Why is every release so anticipated?
Compared to other companies that emphasize safety and regulatory testing, Tesla's strategy is the most controversial and the most eye-catching. Elon Musk has long argued that the key to achieving self-driving is not to stack sensors, but to train a "smart enough brain." This is the background of their choice of vision-only strategy: using a simple camera + neural network combination, relying on a large amount of real driving data and Dojo supercomputers for model training.
But this road has also been plagued by problems. From the FSD Beta controversy, car crashes, to regulatory scrutiny, Tesla is often criticized for "bringing unfinished technology to market too early." Even so, whenever Musk announces the upcoming release of Robotaxi or a new version of FSD, it always attracts a lot of discussion and stock price reactions. On the one hand, this reflects investors' trust in Tesla's self-driving potential, and on the other hand, it also shows that "the self-driving market still lacks a standard answer."
Tesla's approach can be described as a "technology first, rules later" style, trying to pressure regulators to speed up the formulation of new rules through scale and brand influence. This strategy is high-risk and high-return, and it also makes Tesla the most talked-about and challenging role in the entire Robotaxi ecosystem.
Trust and Regulation: Robotaxi’s Biggest Invisible Roadblock
Although Robotaxi technology has gradually been commercialized, the biggest problem is not the technology, but "are people willing to trust it?" After all, letting a car without a driver take you or your family on the road still challenges people's instinct for risk and control.
User acceptance is currently the biggest bottleneck in the expansion of Robotaxi. According to a report released by JD Power at the end of 2024, more than 60% of American respondents said they were "unlikely" to take a driverless taxi, citing reasons such as lack of trust, lack of transparency in information, and inability to respond immediately. Even if Waymo has completed hundreds of thousands of accident-free operations, it is still difficult for people to let down their psychological defenses.
At the same time, regulations are another complex hurdle. Currently, there are still limited states where Robotaxi can be legally put on the road. Even though there are precedents in California and Texas, each state's transportation department has different requirements for data submission, accident responsibility determination, and operating scope restrictions, making cross-state expansion a high administrative challenge. In addition, the insurance system and passenger rights protection framework are not yet mature, increasing the operational risks of enterprises.
From this point of view, the challenges of self-driving cars are actually a bit like the launch of new drugs: the maturity of technology is one thing, and gaining social trust and institutional recognition is another. For Robotaxi to become fully popular, it must pass not only technical tests, but also large-scale social stress tests.
Where is the future of autonomous driving? Infrastructure and platform thinking are the next step
If Robotaxi is to move from the edge of the city to the core of daily transportation, the next key is not only technology, but whether the overall ecosystem is in place. This includes whether the road design is suitable for self-driving cars, whether the vehicle can be integrated with the traffic signal system, and whether the stop point and ride process are intuitive enough.
Taking Phoenix and Austin as examples, local governments have begun to set up exclusive pick-up and drop-off areas and provide V2X (vehicle-to-everything) communication interfaces to help self-driving cars connect to city infrastructure in real time. Technology companies such as Google and Amazon are also trying to connect their own maps and cloud infrastructure to Robotaxi, allowing city managers to track vehicle dynamics and allocate resources.
Another important aspect is platform openness and ecosystem integration. If Robotaxi wants to become a daily application like Uber in the future, it needs to be connected with the ticket booking platform, payment system, operation and dispatch system, etc. In other words, a self-driving car cannot be just a smart car, it must be part of a "transportation platform".
This also means that the next stage of competition will no longer be about whether the model is accurate, but about who can create an open ecosystem that is developer-friendly, government-transparent, and user-intuitive. This is not just an engineering challenge, but also a comprehensive arena for organizational design, business cooperation, and policy negotiation.
Conclusion: What kind of future do we want in terms of mobility?
When Robotaxi becomes a reality, the question we face is not just “can we get on the car?” but “who controls these cars?” Technology may be led by the private sector, but the rules should be jointly decided by society, including data governance, accident responsibility allocation, and fairness in urban transportation.
For entrepreneurs, this is a new track full of opportunities. From sensor modules to data governance, from passenger services to fleet dispatch, every layer is an entrepreneurial opportunity that can be entered. For governments and planning units, how to make new technologies a public interest rather than a regulatory loophole is an important issue in the next decade.
Robotaxi is not the end of self-driving, but a turning point in the future of human mobility. In this transformation, each of us is not only a passenger, but also a decision maker who redefines the relationship between transportation and technology!
Related reports
related articles
Taiwan’s first AI unicorn: What is Appier, with a market value of US$1.38 billion, doing?
What is DNS? Introduction to Domain Name System – System Design 06
Introduction to System Design Components Building Block – System Design 05
Back-of-the-envelope Back-of-the-envelope Calculation – System Design 04
Non-functional features of software design – System Design 03
Application of abstraction in system design – System Design 02
Introduction to Modern System Design - System Design 01
