While Tesla (TSLA) CEO Elon Musk teased details about its robotaxi, called the Cybercab, Amazon’s (AMZN) self-driving subsidiary Zoox has embarked on its own dramatic departure from transportation as we know it.

The difference is clear from aesthetics alone. Zoox's robot taxi has no steering wheel, no driver’s seat, and no pedals. The four seats face inward, as the toaster-like cabin operates bidirectionally, meaning it can drive both ways.

“It’s a vehicle specifically made for people and not for drivers,” said co-founder and chief technical officer Jesse Levinson. “We realized if there isn't a human driver, you can completely change [the riding experience].”

Zoox touts its vehicles as purpose-built, designed specifically for autonomy. Levinson hopes that approach helps propel Zoox to the front of the pack as the company prepares to launch commercial driverless taxi rides in Las Vegas next year.

Zoox's entry into the nascent market would pit it against Google’s (GOOG) Waymo, the market leader in robotaxis so far, and Tesla (TSLA), which offered more hints this week about the evolution of its full self-driving technology.

On Thursday, Tesla CEO Elon Musk unveiled the Cybercab prototype at the company's "We, Robot" event, saying that the EV maker plans to sell the robotaxi for $30,000 and start production before 2027, though Musk's timelines have been delayed before. It also revealed a Robovan concept capable of transporting 20 people.

Before the event, Levinson was quick to point out that the technology Zoox uses is entirely different from that of Tesla. Unlike the Amazon-owned firm, which uses a network of sensors, cameras, and mics, Tesla relies on a camera-based vision system, which is seen as more affordable. Musk has also pushed for the cars to be able to operate in any environment, unlike Google's Waymo, which digitally maps the routes it operates in with a human behind the wheel first before the autonomous cars are deployed.

More consequentially, according to Levinson, Tesla’s autopilot currently uses a driver assistance system, which requires drivers to retake control of the car at any time. That’s fundamentally different from the autonomy-first approach that Zoox is taking, he said.

“It might sound like that's a small difference, but it's actually a giant difference because it means that no matter what's happening in your environment, and it means no matter what's going wrong with your sensors or if they're covered up or if they're occluded, or even if they fail, you still have to be able to keep driving safely,” he said. “Teslas and passenger cars today, they're just not designed that way. And so it's a really, a very different hardware and software product entirely.”

Zoox's robotaxi model

Autonomous cars remain an intriguing ambition for tech companies, despite delays in the wider rollout of driverless taxi services. Consulting firm McKinsey & Co. estimates the global robotaxi market could generate as much as $1.3 trillion in revenue by 2030.

But implementation has been fragmented, with commercial operations limited to a handful of cities. Companies that have successfully made it to market have opted to retrofit cars intended for drivers to achieve full autonomy.

Zoox’s approach pivots away from that model. The company has specifically relied on a sensor suite that includes lidar, radar, and cameras to navigate a limited stretch of roads. The vehicles, built in-house, include a sound bar surrounding the cabin, with mics built into the interior and exterior of the vehicle, in part to enhance the detection of emergency vehicle sirens.

While Zoox assembles vehicles at its own manufacturing facility in California, the company purchases large modules from suppliers to control overall costs tied to automotive manufacturing, according to the firm.

Levinson said vehicles are equipped with redundancies for critical components to ensure safety. There are two braking systems, two steering racks, two batteries, two motors, and four sensor pods in each corner.

"If any one of those components has a hardware or software failure, not only is it still safe, but we can actually keep driving and take the customer to their destination," Levinson said. "Just like if there's a component failure on an airplane, you still land the plane, you don't drop out of the sky."

Those redundancies are coupled with human technicians, who monitor the vehicles for safety. While the cars operate fully autonomously, technicians step in on rare occasions to guide the vehicle when it encounters unfamiliar situations, like construction zones, according to Levinson.

"The vehicles are only getting help from humans only 1% of the time," Levinson said. "As it gets more sophisticated, the amount of time they’re asking for help goes down and down and down and down. We don’t think it ever truly gets to zero, but it doesn’t have to get to zero. It just has to go down to something that’s very small."

For now, Levinson said he’s confident that Zoox is on the right track for the future. It is on target to expand beyond its 5-mile test loop in Las Vegas, with commercial service set to launch in 2025.

“In the long term, more and more companies will converge on something like this,” he said, pointing to his vehicle, “where you've really imagined the experience without needing all the legacy of a car.”

Click here for the latest stock market news and in-depth analysis, including events that move stocks

Read the latest financial and business news from Yahoo Finance