 |
| DoorDash is testing autonomous delivery using robots that travel on sidewalks, but CEO Tony Xu said using robotaxis would require an "end-to-end system." REUTERS/Carlo Allegri |
While self-driving cars are starting to revolutionize ride-hailing services in select U.S. cities, don’t expect a robotaxi to drop off your next burger and fries anytime soon at least, not according to Tony Xu, CEO of DoorDash.
Speaking during the company’s earnings call on Wednesday, Xu explained why autonomous food delivery remains far more complicated than simply ferrying a passenger across town. Although Uber and Tesla have introduced limited self-driving ride services in cities like Austin, San Francisco, and Atlanta, Xu cautioned that applying that same technology to food delivery is an entirely different challenge.
“The passenger can walk in and walk out of the car, even if the drop-off or pickup locations aren’t perfect,” Xu said. “Deliveries, by contrast, require a much more precise hand-off between the restaurant and the vehicle.”
In other words, a person can accommodate a few quirks in routing or parking but your pad thai can’t.
Autonomous Delivery Needs More Than a Self-Driving Car
Xu described the food delivery process as one that requires an “end-to-end system” not just an autonomous vehicle, but a seamless connection between restaurant preparation, pickup, secure transport, and precise customer drop-off. This complex chain of tasks introduces variables that robotaxis aren’t yet equipped to manage.
“That’s probably the single biggest learning we’ve had,” Xu said of the company’s experiments with autonomous delivery.
Those experiments, however, are far from stagnant. Xu pointed to positive results from DoorDash’s partnerships with Coco Robotics, a startup that develops small wheeled delivery robots designed to traverse sidewalks. The companies are currently testing these bots in Chicago and Los Angeles, with previous pilot runs conducted in Finland under Wolt, DoorDash’s international subsidiary.
Xu remained optimistic: “Autonomous delivery is something we’re very excited about,” he said, while acknowledging it’s still in the experimental phase.
Robotaxis May Be Giving Rides But Not Delivering Dinner
While food delivery is still figuring out the AV puzzle, robotaxis are already on the road at least for passengers.
In June, Tesla rolled out a limited version of its robotaxi program in Austin, with safety employees seated in the vehicle during rides. The company has since expanded to San Francisco, where those safety personnel now sit in the driver’s seat rather than the passenger side. Similarly, Uber is partnering with Waymo to provide fully autonomous rides in Atlanta and Austin, and plans to bring more self-driving cars into its network via collaborations with EV maker Lucid and autonomous tech firm Nuro.
But while giving a human a ride may require a functioning AV platform and good maps, getting a taco to your front door while it’s still hot and without a person onboard is a whole other engineering feat.
Smaller Deliveries, Smarter Vehicles
For Xu and DoorDash, the future of automation likely doesn’t lie in large, car-sized robotaxis, but in smaller, more efficient autonomous solutions.
“You don’t need a 4,000-pound vehicle to deliver a one- or two-pound item or package,” Xu remarked during an earlier earnings call in May.
This sentiment aligns with DoorDash’s ongoing exploration of lightweight, ground-based delivery robots that could navigate sidewalks and urban environments without contributing to road congestion or parking challenges. These units may ultimately offer a better fit for food and grocery delivery, rather than trying to retrofit existing robotaxi infrastructure for those needs.
In short, while autonomous vehicles are showing promise in the ride-hailing world, Xu believes the food delivery sector requires a more customized, nuanced approach one that’s still under development.
