DoorDash Delivery Robots Reveal a Key Shortcoming of Automation by Still Relying on Human Drivers to Load Orders
DoorDash has built a delivery robot capable of navigating roads and sidewalks, but the machine still needs a human being to carry the food a few feet from the restaurant to the curb.
The company is offering some gig workers about $5 to drive to participating restaurants, collect customers’ orders and place them inside Dot, DoorDash’s red-and-white autonomous delivery robot.
The unusual assignment is being tested in the Phoenix area, where DoorDash has begun using Dot to complete select deliveries. The arrangement offers a glimpse at one of automation’s persistent shortcomings: Even sophisticated machines can struggle with the ordinary physical tasks that people perform without much thought.
Drivers Paid to Complete the Robot Handoff
Devena Bybee, a DoorDash driver in Mesa, Arizona, told Business Insider that she received an offer for one of the robot-loading assignments in early July.
Drivers participating in the program said the jobs pay approximately $5 and can take about five minutes. Workers travel to the restaurant, retrieve the food, place it inside Dot and photograph the process to show that each step was completed.
Once loaded, the robot handles the longer portion of the delivery.
Dot uses lidar, radar and cameras to travel along roads, bike lanes and sidewalks at speeds of up to 20 mph. However, it cannot enter a restaurant, pick up a bag from the counter or load the food into its storage compartment.
That leaves DoorDash drivers responsible for what may be the shortest — but most physically complicated — part of the trip.
DoorDash described the program as a limited pilot designed to help restaurants manage orders during busy periods. The company said human drivers will continue completing most deliveries.
Automation Still Depends on Human Labor
DoorDash is not the only technology company relying on workers to handle small tasks its automated systems cannot complete.
Waymo has reportedly paid people to close doors that passengers leave open on its autonomous taxis. Other robotics companies have used remote operators or on-site workers to help machines deal with unexpected obstacles and situations.
The examples underscore a continuing challenge for autonomous technology. A robot may be able to map an entire neighborhood, recognize traffic and travel several miles, yet still be unable to grab a restaurant bag or close a car door.
Also Read: Side Hustles Become a Necessity for Millions as Costs Rise
DoorDash has continued investing in several forms of automated delivery. The company has also received Federal Aviation Administration clearance connected to its drone-delivery operations. Still, moving an order between a restaurant employee and an autonomous vehicle remains a difficult part of the process.
DoorDash Delivery Robots Add Another Gig-Economy Job
The pilot creates a new and somewhat ironic layer of gig work: People are being paid to prepare robots to perform work that previously belonged entirely to people.
For DoorDash, the arrangement allows Dot to remain in service without requiring restaurants to send employees outside to load it. For drivers, it offers a quick payment without requiring them to complete the delivery themselves.
But the model also shows that automation does not always eliminate human labor. In some cases, it breaks one job into smaller assignments and places workers at the points where machines cannot reliably function.
The biggest challenge in urban robotics may not be developing faster artificial intelligence or adding more sensors. It may be handling the unpredictable, physical handoffs that occur in restaurants, homes and public spaces.
For now, DoorDash’s robot can make the trip. It just needs a person to pack its bag first.
