Start with the sensors, then design the rest: How Zoox built its robotaxi

In the ever-evolving landscape of transportation, the focus on artificial intelligence and robotics is stronger than ever. Nearly a decade ago, the buzz was all about self-driving technology, with robotaxis dominating discussions at industry events. While the surge of startups in this space has dwindled post-COVID, significant advancements in the technology have emerged. Cities like San Francisco and Austin are now witnessing a fleet of sensor-equipped vehicles navigating their streets. Among these, the distinctive pod-like robotaxis from Zoox, a company acquired by Amazon in 2020, are making waves. Unlike other developers who retrofit traditional vehicles like the Hyundai Ioniq 5 for autonomous driving, Zoox is unveiling a purpose-built model designed specifically for ride-hailing services, currently operational in Las Vegas and San Francisco. Chris Stoffel, the director of robot industrial design and studio engineering at Zoox, emphasizes the unique requirements of a robotaxi compared to human-driven cars. The design begins with strategically placed sensors atop each corner of the vehicle, providing an unobstructed, comprehensive view essential for navigating urban environments. Ryan McMichaels, the director of sensor engineering, notes that this setup allows for superior frontal coverage, a feat that is challenging to achieve with modified vehicles. The symmetrical, bidirectional design of the Zoox robotaxi also sets it apart. This innovative approach eliminates the need for cumbersome maneuvers like three-point turns, enhancing the vehicle's agility. With both axles capable of the same steering degrees, the Zoox robotaxi can maneuver in tight spaces more effectively than competitors like the GM Hummer EV, positioning itself as a versatile option for on-demand urban transport.