Earlier this year six Amazon Scout delivery robots rolled out in a pilot program in Snohomish County, Wash. The boxy bots, which resemble six-wheeled ice chests, carry meals, groceries and packages to homes and offices in this region just north of Seattle. They join a small-but-growing number of automated couriers trundling down the sidewalks of London, Beijing and other cities and communities worldwide. These machines must run a gauntlet of pedestrian legs, nosy dogs and cracked pavement. Which raises the question: Why are companies investing in delivery bots at all?
These services are gaining traction as a growing number of city dwellers expect immediate or scheduled delivery for just about everything. Between 2016 and 2017 online retail sales in the U.S. increased by 16 percent. Meal-kit companies are proliferating and grocery stores are making an increasing percentage of their sales online. On the final leg of all these deliveries, called the last mile, humans on bicycles, motorized scooters or large delivery trucks typically ferry packages. All those vehicles compete for space on busy urban streets. “Deliveries are trending upwards in all dense city centers, and if city and state leaders don’t start thinking about innovations like robot deliveries, we can expect even worse traffic jams when we’re all trying to get places,” says Paul Mackie, director of research and communications at the Mobility Lab, a transportation policy research center in Arlington, Va.
A study by Mobility Lab and George Mason University found 73 percent of freight and delivery vehicles in Arlington were parked outside of authorized areas, often blocking bike lanes, fire hydrants and crosswalks. By moving the last leg of deliveries from the road to the sidewalk, cities could reduce congestion and eliminate the parking problem entirely, Mackie says.
Entities such as Amazon are not developing this delivery technology simply to clear up urban traffic. Self-driving vehicles and sidewalk robots could slash last-mile delivery costs in cities by as much as 40 percent, according to a 2018 report by the New York–based McKinsey & Company consultancy firm. A delivery robot can cost thousands of dollars to manufacture, and most currently require human supervision and maintenance. But in the long run companies that stake a claim to autonomous delivery vehicles in the next several years could end up saving billions of dollars, the report stated.
To gain public trust, these machines must demonstrate they can safely and unobtrusively share pedestrian spaces. Some U.S. cities have fairly empty streets and sidewalks, which might not see a pedestrian pass for long stretches of time. These paths could accommodate robots, says Renia Ehrenfeucht, chair of the Community and Regional Planning Department at The University of New Mexico in Albuquerque and co-author of the book Sidewalks: Conflict and Negotiation in Public Space. When the pavement gets more crowded, however, even robots rolling along at walking speeds will face challenges, which will get worse in U.S. cities with narrow sidewalks. “It’s actually really hard to navigate crowded sidewalks and not bump into people, and do it smoothly,” Ehrenfeucht says. “Until delivery robots are that skilled, if they could be, they will be disruptive.”
Sidewalk robots may also face disruptions to their own routines, including curious bystanders blocking their paths and thieves looking to snatch a package en route. The latter problem could be solved with remote-controllable locking mechanisms on packages. Still, one video shows a delivery robot being temporarily stranded after kids covered its cameras and sensors with snow. Another cautionary tale comes from the Canadian hitchBOT social experiment, in which a simple humanoid robot traveled across Canada and Europe—before ending up decapitated in Philadelphia at the start of its U.S. journey. “I do think the largest problem is going to be either prankster or outright malevolent behavior toward these things,” says Mary “Missy” Cummings, director of the Humans and Autonomy Lab at Duke University.
Starship Technologies, a company based in San Francisco and Estonia, takes a more optimistic view. Hundreds of their robots have already made 25,000 deliveries, navigating local neighborhoods using machine-learning software, onboard sensors and digital maps. The majority of people who share the sidewalk with Starship’s robots hardly notice them, and most attention seems positive, says Ryan Tuohy, Starship’s senior vice president of business development. “After our robots have been in an area for a while, people get used to them. They become part of the community and residents look out for their well-being,” Tuohy says. Starship’s robots operate almost entirely autonomously in mapped areas, but remote human operators monitor them in case they need to intervene. Still, even Starship previously admitted people have occasionally given its $5,500 robots a kick in passing.
Additional challenges include perfecting the software that helps delivery robots avoid both still and moving objects as well as dealing with city officials who want to protect public spaces. For example, San Francisco, home to delivery start-ups such as DoorDash and Postmates, slapped a ban in 2017 on autonomous deliveries by sidewalk robots and has only slowly opened the door for companies seeking permits. Other U.S. communities have regulated where such robots can go and how they must behave.
Given these obstacles, sidewalk delivery robots are not necessarily destined to win the future. Nuro, a start-up founded by two ex-Google employees, has developed driverless “R1” delivery vehicles that cruise on streets. The R1 cars join Nuro’s preexisting fleet of self-driving Toyota Priuses that, as of mid-December 2018, had already made about 1,000 grocery delivery runs in Scottsdale, Ariz. As for flying delivery drones, they face their own challenges. But the technology could offer more widespread services in rural and suburban areas that lack well-maintained sidewalks.
Ultimately, the best solution to a glut of delivery vehicles could be less high-tech. A 2016 McKinsey report pointed out old-fashioned bicycles remain the most cost-competitive choice for many last-mile deliveries. “If droids do not become significantly cheaper,” the analysts wrote, “bike couriers are likely to be the best delivery form for instant delivery in urban areas.”