By Tom Gresham
It was at the Urban Freight Lab at the University of Washington that supply chain experts coined the increasingly used term “the last 50 feet of delivery,” according to Anne Goodchild, professor of civil and environmental engineering and founding director of the Supply Chain Transportation & Logistics Center at Washington (where the Urban Freight Lab is housed). Goodchild said that the final 50 feet represents a short but critical segment of the delivery journey—a distinctive piece of the process that presents challenges and opportunities that are unique among the various components of the supply chain.
The last 50 feet also is a portion of delivery that until recently had been largely overlooked, Goodchild said. “We haven’t optimized for this part of the supply chain yet,” she said. However, “now that’s where the attention has really turned.”
With attention comes an assortment of innovative ideas for improving efficiency and effectiveness, including a range of technology-based solutions. Among those is the use of robotics to help complete the delivery process by transporting a product from its delivery vehicle to its final destination. Automation has been applied in areas such as package sorting, route planning and schedule optimization in the delivery process, but until now it has not yet made a major impact in the actual delivery of packages. The type of robotics used in the potential delivery solutions under development vary, encompassing such styles as drones, wheeled robots and legged robots. In some of the scenarios under consideration, robotics are paired with automated, self-driving vehicles.
Monica Eaton-Cardone, COO of Chargebacks911, which manages more than 2 billion transactions annually for online merchants, said robotics and other ambitious attempts to find efficiencies in the final 50 feet of delivery can be traced to the intense, e-commerce-fueled demands on supply chain companies to satisfy the expectations of “a more entitled consumer.” The push to ensure goods are delivered with more frequency and more speed is putting strain on every leg of deliveries.
A robotics solution that works—one that is able to navigate the many challenges associated with using robots in deliveries in public—would be a major boost to effectiveness in the final 50 feet, potentially able to provide precision, reliability, transparency and competitive advantage to the companies that employ it, Eaton-Cardone said.
“When you consider changing consumer demands, robotics satisfies those demands almost impeccably,” Eaton-Cardone said. “There’s no way to achieve the necessary scale and efficiency needed for deliveries without technology.”
Still, Eaton-Cardone said many questions will need to be answered before robotics becomes a widespread solution to the last leg of delivery.
“I do think that there’s more upside than downside,” Eaton-Cardone said. “But we still have a lot to do to figure out what needs to be done to reduce the risk that comes with it.”
Robots have recently made new inroads into delivery, crossing the boundary between the conceptual and the actual. For instance, Amazon has introduced wheeled robots for deliveries in limited deployments, and Starship Technologies announced a plan to introduce thousands of its wheeled delivery robots on college campuses. In another case, Agility Robotics is working with Ford to combine Ford’s autonomous vehicle technology with Agility’s bipedal robot Digit for an automated delivery process. As Damion Shelton, CEO of MODEX 2020 exhibitor Agility Robotics, noted, the use of robots for the final 50 feet of delivery is not speculation about a distant, abstract future—Agility Robotics will be managing its first deployment of Digit robots to Ford and others in the first quarter of 2020.
Still, that does not mean that robots are on the verge of becoming a common sight delivering packages to front doors. Shelton said he believes the integration of robotics into delivery will be more gradual than the arrival of cell phones, for instance. “The cell phone rollout occurred and was almost completed in about 10 years when virtually everyone had a smartphone.”
“It’s going to be surprising and unusual for a while to see robots,” Shelton said.
The final 50 feet
Goodchild said the Urban Freight Lab selected two key metrics to emphasize in measuring efficiency and effectiveness in the final 50 feet of delivery: dwell time, which is the amount of time a delivery vehicle stays parked while a product is delivered from the vehicle to the recipient, and failed deliveries. Reducing both dwell time and failed deliveries are beneficial for communities, companies and consumers, she said.
Goodchild said decreasing dwell time allows each driver to complete more deliveries on their route, aiding the bottom line for businesses, while reducing the number of vehicles needed to make those deliveries to ease the strain on a community’s roadways. Similarly, reducing failed deliveries, which can include theft, limits wasteful repeat trips, Goodchild said.
Goodchild said the Urban Freight Lab and others are investigating a wide array of solutions to the final 50 feet. The use of robotics will be challenging, she said. So far, Goodchild said, robotics solutions have been most successful in manufacturing environments where there are specific, limited tasks to complete and the space is controlled and organized for the machines. Ultimately, she said the biggest challenge to integrating robotics into the last 50 feet of delivery will be operating in an uncontrollable space and sharing it with the rest of the world.
Shelton said the variety of delivery environments means that the last 50 feet is where the bulk of human labor currently occurs.
“In general, there’s not a lot of commonality,” he said. “Someone who lives in an apartment building is going to have a very different set of delivery requirements than I am, living in a more suburban neighborhood trending to the rural… It’s just an incredibly varied environment. It represents the least amount of structure in the history of automation.”
Preparing a robot for the unexpected encounters it would face on deliveries represents a much stiffer challenge for automation solutions than the auto-pilot feature on a plane or even a self-driving vehicle in a controlled environment, Shelton said.
“The number of times when something weird happens initially is going to be quite high because things that you and I take for granted—like, say, a little kid riding a tricycle—maybe the robot hasn’t seen before,” Shelton said. “From a safety standpoint, you’d better not hurt people when you’re doing these kinds of tasks. I think that is the single biggest technical barrier to pulling this task off. It’s about making sure that you’re able to handle the real world. Because it gets quite complex and ground-based locomotion in unstructured environments represents the hardest possible thing to be doing. But it’s also why it’s a large opportunity for us.”
Shelton said Agility Robotics’ focus on legged robots stems in part from the unpredictability of environments and wanting to equip the robot with similar mobility to people.
“The first thing is, ‘Technologically, can you even automate this?’” Shelton said. “Secondarily, the question is, ‘Is it profitable to do it? From a business perspective, if you can technologically pull off the automation, is it cost effective to do it?’ For some applications, the answer’s going to be yes on shorter time frames. And others for very complicated delivery tasks, even if you have one of our robots and even if it’s working perfectly, it’s probably still going to be too complicated in some percentage of environments for it to be economically feasible to pull it off.”
However, even for those delivery environments that prove too challenging in the near future, Shelton said “it’s going to improve over time. Just like anything else. For instance, early on you didn’t find autopilots in low-cost aircraft. Now they’re pervasive.”
Acceptance and public policy
Goodchild and Eaton-Cardone noted that replacing human delivery professionals with robots will depend on robots becoming at least as effective and reliable as the humans, and Shelton said a key piece of making robots acceptable to consumers will be ensuring that the use of the technology does not worsen the customer experience.
“There’s a customer acceptance aspect, which means that part of making it economically feasible is to make people want to use the service,” Shelton said. “And the easiest way to do that is to give them an experience with it that is similar or identical to what they already have.”
Shelton said he does not believe that the use of robots in delivery ultimately will lead to a loss of jobs. He points to the airline industry’s increased use of automation as an example, which he said led to less expensive airline travel and increased demand—meaning the number of jobs increased.
“I expect something very similar is going to happen with delivery,” Shelton said. “You’re going to have people who are doing progressively harder, more interesting jobs in that space where even if you could prune off the bottom half of the workload, the workload has gone up by a factor of three. So there’s still a net gain in employment.”
The involvement of robotics in delivery does not just depend on solving a range of technical and economic questions, Goodchild said.
“There are some pretty substantial sort of public policy questions that we’ll need to get our hands around before there’s really a reliable market and people are going to invest in this heavily,” she said.
Eaton-Cardone agreed. For instance, she said the use of robotics in delivery begs questions related to the regulatory environment for the devices. Will they face inspections like vehicles? What should insurance requirements be? She said the use of robotics will mean that security safeguards to prevent hacking of the machines and other criminal exploitation will be essential.
“It will be important to make sure we do a good job reducing risk in the policy arena,” Eaton-Cardone said.
Goodchild said it’s inevitable that more sophisticated technology will become an increasingly large factor in the final 50 feet of delivery. The extent that it will involve robotics delivering products to front doors remains an open question.
“It’s exciting,” Goodchild said. “They have to be tried and they have to be tested, but from an engineering and just sort of human development perspective it’s exciting to see—but also to start to realize why it’s so hard to do.”