Arriving in a sleek, silver two-door 'Cybercab' on the fake streets of the Warner Bros. movie studio set, Elon Musk unveiled Tesla’s vision for a fleet of fully autonomous cars, along with their futuristic-looking 'Robovan.' However, disappointment came from the lack of concrete timelines, limited technical details on its rollout, seemingly optimistic price targets, and the absence of a fully operational real-world demonstration. Despite the somewhat underwhelming presentation, Tesla's robotaxi plans have sparked debate about which companies are truly poised to capitalize on this technological revolution.

The development of autonomous driving technologies relies on a few core technologies that are each crucial to enabling vehicles to navigate their environment entirely without human intervention. These crucial technologies include sensors, artificial intelligence, and high-definition mapping.

Sensors, which provide the vehicles with real-time data about their surroundings, are central to autonomous driving. The three main types used are LiDAR, radar, and cameras. LiDAR creates highly accurate 3D maps of the surroundings with the use of laser beams, which allows the vehicles to detect objects at long distances under various lighting conditions. The precision of LiDAR is unmatched, an advantage that also comes with a heavy price tag. However, the technology isn’t bulletproof and can still struggle in adverse weather like a heavy rainstorm or fog. While radar is more affordable and functions reliably in all weather conditions, its resolution is lower, making it harder to distinguish fine details. Cameras on the other hand provide high resolution images that are crucial for identifying visual cues like lane markings and traffic signs.

While they are cost-effective, they also depend on lighting conditions, which makes them less reliable in low light environments or adverse weather.

Source: Eetimes

Each player in this space has chosen to adopt a different combination of these technologies. Waymo, a subsidiary of Alphabet, is widely regarded as the current leader in this field with its sensor-heavy approach that integrates LiDAR, radar, and cameras to provide several layers of data. Implementing this variety of sensors ensures the vehicles can operate safely even when one sensor becomes unreliable. This approach, however, comes with significantly higher costs. The expensive LiDAR systems might give them a head start in this race, but they also make it challenging to scale this technology to mass-market vehicles with an affordable price tag. The company has already launched fully autonomous Robotaxis in a handful of cities like Phoenix, LA, and San Francisco.

Another major player in the US is Cruise, the autonomous driving subsidiary to General Motors, which has opted for a similar path as Waymo, leveraging all three sensor technologies to emphasize safety.

Tesla, in true Tesla fashion, has decided to take a radically different approach. The company has chosen to forgo LiDAR altogether and instead rely on a vision-based system that uses cameras as its primary sensors, complemented by radar. The company’s full self-driving (FSD) system uses AI to process the visual data in real time and make driving decisions based on what the cameras see. This approach is far more cost-effective than that of its rivals as cameras are significantly cheaper than LiDAR systems. Another advantage for Tesla is that the company benefits from already having millions of vehicles on the road equipped with FSD hardware, allowing them to collect huge amounts of driving data, which is a key asset in refining its AI systems.

However, Tesla’s cost-effective approach is not without its setbacks. As previously mentioned, cameras are highly dependent on lighting and weather conditions. Additionally, lacking the depth perception offered by LiDAR means that Tesla’s system may be less reliable in handling complex scenarios such a crowded intersections or obstacles in low visibility. Their belief is that their AI systems will eventually overcome these limitations through continuous learning and remote software updates.

Furthermore, in addition to all the on-board sensors, connectivity will play a key role in enhancing the cars’ capabilities through IoT integration. Connected vehicles will be able to communicate with each other as well as with infrastructure like traffic lights and share real-time information on traffic patterns or roads hazards. This connected network is often referred to as “vehicle-to-everything” communication. Players like Google who already have extensive mapping data and IoT expertise will likely have a significant advantage in building a highly efficient robotaxi ecosystem.

While technology is the engine powering the Robotaxi revolution, it’s the regulatory environment that acts as either a gatekeeper or an accelerator. In the U.S. for example, regulations may vary significantly between states, which is something that companies must navigate. States like California and Arizona, where Waymo operates, have been fairly open to the testing of driverless vehicles on public roads. Understandably, stringent safety requirements remain in place, meaning that a full deployment of this technology hinges on the companies being able to prove that their vehicles can match or exceed the safety standards. In Europe, unsurprisingly, regulators are taking a more cautious approach, prioritizing strict safety guidelines and data privacy. The European Union has laid the groundwork for autonomous vehicle regulation, but individual countries retain some autonomy in deciding how fast Robotaxis can be introduced, creating a slower, more fragmented and uncertain market compared to the U.S.

Faced with intense competition in the Robotaxi race, Tesla has different key scenarios at its disposal.

Scenario 1: Tesla as a car-selling company

Tesla focuses solely on selling AVs directly to end-consumers. This path would position Tesla as a car manufacturer primarily, similar to its current model, but with AV-specific offerings. Individuals or businesses could purchase Tesla's self-driving vehicles for personal or commercial use. Tesla could profit by selling the hardware, then charging a subscription for software updates like Full Self-Driving (FSD) to enable full autonomy.

Scenario 2: Partnerships with ride-hailing companies

Tesla could partner with established ride-hailing platforms like Uber, Lyft, or Bolt, which already have extensive networks and customer bases, Uber, for example, has around 150 million monthly active users globally. By selling its Cybercab Robotaxi to these ride-hailing companies, Tesla could tap into this massive network while focusing on hardware and software, leaving customer service to the platforms. There are signs of a growing Tesla-Uber partnership, such as Tesla offering up to $3000 discounts to Uber drivers purchasing its vehicles.

Uber's CEO, Khosrowshahi, has expressed in the past interest in collaborating, telling the Financial Times that he’d "love to have [the Cybercab] on the platform," Khosrowshahi highlighted Uber's extensive experience, stating, “It’s taken us 15 years. It’s taken us tens of billions of dollars of capital, and we can provide that instantly to a partner. Hopefully, Tesla will be one of those partners.” In this model, Tesla could maintain long-term involvement through vehicle maintenance, offering subscriptions for FSD software, or taking commissions on each ride where a Tesla Cybercab is used.