Innovation
Blogs •
Boston Dynamics is testing a last-mile delivery solution where Spot will help get packages from van to doorstep with less strain and greater efficiency.
Today, it’s a demo: Spot®, the quadruped robot from Boston Dynamics, hops from the back of a delivery vehicle. A driver loads two boxes onto a conveyor payload on the robot, and then Spot trots through the neighborhood. It makes its way up walkways, crouches low to the ground, and gently delivers each package directly to the right doorstep before returning to the van.
Soon, Spot may help delivery drivers move more packages, more quickly, through some of the most expensive and physically demanding stretches of the logistics chain. Eventually, autonomous vehicles could even transport entire fleets of Spot delivery robots in some areas.
In between, engineers must solve countless challenges to not only bring autonomous last-mile delivery to life, but also to put Spot to work at scale in public, largely unstructured settings.
“So much of logistics is already automated, but we believe that the final frontier of logistics automation is that last 50 feet,” says Marco da Silva, Vice President and General Manager for Spot at Boston Dynamics. “We’ve done things to automate inside of the hub, inside of sort facilities, but now we’re trying to bridge the gap to the final mile.”
Since Spot first became commercially available, the robot has mostly been completing highly repeatable actions in industrial environments. For example, many companies use Spot to conduct routine inspections and data gathering at sites like factories and energy production facilities.
The future of truly general purpose robotics—robots that can go anywhere and accomplish any task—requires a much higher bar of adaptability and safety than industrial automation. For example, a robot with a service job has to interact predictably around the general public, not only trained professionals. Boston Dynamics has the engineering capacity to solve any number of problems, but to prioritize new applications, they look for opportunities that both target the hardest technical hurdles and provide real value over existing solutions.
Last-mile delivery presents exactly that kind of opportunity. According to some estimates, last-mile delivery now accounts for more than half of total supply chain costs, with U.S. carriers spending around $90 billion per year on last-mile package delivery. Currently, almost all of this work rests on the literal backs of the drivers, who are carrying packages down driveways and up stairs for hours a day. The challenge of the last mile gets increasingly difficult in just the last 50 feet of a delivery.
By training Spot to assist with these tasks, Boston Dynamics leaders hope drivers will be able to deliver more packages more quickly, while also reducing the physical strain on drivers.
The unstructured environments of suburban neighborhoods present myriad little obstacles to overcome or avoid: snow and ice, dog walkers and joggers, kids on bikes, walkways blocked by trash bins, lawn sprinklers whose spray spills over onto sidewalks, and streets where no two porch designs are the same.
For Paige Miller, senior staff product manager for Spot, that challenge is the ultimate test and the place where past last mile automation has stalled out. Robots cannot assume clear paths, flat surfaces, or standard drop-off settings. Instead, they must navigate curbs, stairs, gravel, and other variations that would be smoothed out and standardized in a more controlled environment.
“Companies have tried drones for delivery, but they haven’t been able to get to the porch, and so packages were being left in the middle of the yard,” Miller explains. “We’ve also seen the struggles of various wheeled robots for last-mile delivery. So, I think the industry is looking for something in this space where other robots have fallen short of human parity. Spot’s legs and ability to handle unstructured terrain and environments give it advantages over other robots.”
In other words: The challenges that make last-mile delivery essentially impossible for other robots make it an ideal use case for an all-terrain, legged robot like Spot.
Boston Dynamics’ last-mile demo video shows a driver loading two packages onto Spot, and it shows the robot delivering the packages to two different houses. However, the video doesn’t show what drivers will be doing while Spot is off making deliveries: either staging the next set of packages or else delivering other packages to other houses and then meeting up with Spot again at the vehicle for the next stop.
Eventually, Miller says, delivery vehicles may even be outfitted with ceiling-mounted robotic arms that automatically stage packages for Spot. But at first, the focus will be on simply providing delivery drivers with a robotic partner that helps them get more done, with less physical effort.
“For every three packages delivered via Spot, we think we can add another package to the van,” Miller says. “Spot can carry two 16x12x10-inch parcels, and we think that will capture at least 60 percent of the average van’s packages.”
The goal is not just to have Spot make deliveries in suburban neighborhoods (where the robot’s last-mile efforts will likely start out), but to do so with the same level of care and attention to detail as delivery drivers. The conveyor on top of the robot will be outfitted with stop sensors, making it possible for Spot to split a single payload into multiple doorstep deliveries, as well as a small tray that actuates forward to reduce drop speed. “We’ve already done a lot of work on the payload prototype to make sure the delivery is soft,” says Miller. “We even delivered a carton of eggs to make sure we were demonstrating the necessary level of package care.”
That standard of care also means making sure that Spot doesn’t trample flower beds or stomp across manicured lawns. Initially, deliveries may be made using driver control. Then, these routes can be saved, allowing Spot to make future deliveries autonomously to known addresses. This is essentially the same process already followed in industry, where operators first manually walk Spot through inspection routes to create maps for autonomous missions.
Boston Dynamics is actively seeking logistics partners to help work out the details of how last mile automation scales. Miller and her team are already in talks with major logistics companies about testing Spot for a last-mile delivery solution. These early partners will help to validate and refine the application, and also provide input on what is needed to make it work in the real world.
The goal is to move beyond limited demos to a full pilot project with an industry leader: Spot working in concert with the driver to deliver 200 packages a day, five days a week. Lessons from that pilot will inform a scaled-out launch.
In between, Boston Dynamics engineers will likely have to solve just as many challenges as Spot will face on its 50-foot delivery missions. In addition to helping Spot overcome physical obstacles, they will need to find ways to facilitate the partnership between drivers and robots, optimize routes, and ensure that Spot is able to accommodate customer preferences, building on tools like Orbit™, Boston Dynamics’ orchestration and intelligence software. At least initially, Miller says, teleoperators may be on standby to ensure safety and intervene if Spot gets stuck, similar to how autonomous taxi companies operate today.
The most valuable answers may result from questions that aren’t even being asked yet. Those lessons, Miller says, will come not from the lab, but from the field. “If we disappeared into the lab for five years and came out with a solution, we would miss the mark in a variety of ways,” she says. “That’s why we need partners. We don’t want to do this in a bubble.”
Recent Blogs
•7 min read
Stretch Enhances Logistics & Maintenance at Otto Group
•9 min read
A Retrospective on Uses of Boston Dynamics’ Spot Robot
•43 min watch
The Art Behind Human-Robot Interaction
Have a question about our robots and applications? We're here to help. Reach out to our sales team to learn how advanced mobile robots can impact your operations.