USC Viterbi Research to Streamline our Robot Delivery Future

| June 22, 2022

A new study into optimized symbiotic vehicles for use in warehouses and for robot deliveries has been funded by Toyota’s University Research Program.

Delivery robots are being trialed in cities in the U.S. and around the world. But most are controlled by a remote human pilot. Image/VCU Capital News

Delivery robots are being trialed in cities in the U.S. and around the world, but most are controlled by a remote human pilot. Image/VCU Capital News

John Carlsson can’t wait for a future where robots roam the streets, seamlessly darting back and forth from larger delivery vehicles to bring us our food, goods, mail, and medicine at lightning speed. A future where automation can help make the world more accessible for people with mobility issues, with fewer vehicles clogging our roads on time-consuming errands that add to the carbon footprints of cities.

Carlsson, the Kellner Family Early Career Chair and associate professor of industrial and systems engineering, has just been awarded $200,000 to develop optimization techniques to help make this bold new future possible, with Toyota’s Raymond Corporation announcing him as one of just three recipients in their University Research Program. Over the next year, he will design a framework to show how fleets of “helper vehicles” such as automated ground-based robots or aerial drones can work quickly and effectively in tandem with a larger delivery vehicle in settings such as automated warehouses, or for deliveries in urban centers.

It’s a system that he calls “symbiotic vehicle routing.”

“Imagine in Africa we see a hippo with a bird sitting on it, picking off food and cleaning its teeth — there’s a big lumbering thing, and then a tiny, little thing helping it out. They’re both benefiting from each other,” Carlsson said. “Symbiosis is a good biological metaphor for what we’re doing — showing how big things are good at some things and small things are good at others.”

Carlsson and his team will first examine how this can be applied in a warehouse setting, an area where he has previously worked on optimization solutions to allow automated robots to navigate effectively around the aisles of massive unmanned warehouses. Carlsson’s prior research led to his induction as an Edelman Laureate by the Institute for Operations Research and the Management Sciences (INFORMS).

“There have been a lot of cool developments in warehouse research in the last 20 years because we have AI that can keep track of where everything is, and there is so much data available. This offers a whole new way of thinking about how you can move things around,” Carlsson said.

John Carlsson

Kellner Family Early Career Chair and Associate Professor of Industrial and Systems Engineering John Carlsson.

“In a symbiotic system, you could have a big forklift driving around alongside little robots on the ground that carrying individual items and dropping them where they need to go,” he said.

Carlsson and his research team will be using an algorithm to enable the symbiotic routing system in which a large host vehicle in a warehouse can move an entire rack shelf while being assisted by a fleet of automated guided vehicles (AGV) “helpers” that pick individual items. The algorithm, which has already been designed, will be harnessed by the team in real-world experiments to find the most effective routes between the host and the helpers.

The symbiotic approach would speed up the processing of goods and free up storage space, offering customers the potential of same-day delivery of goods, which could be useful for urgent deliveries of medication, such as insulin.

“If the big vehicle is capable of handing things off to the small thing, we want to know how much we can benefit — the dollar amount, the time amount,” Carlsson said. “Can we go from overnight to same-day delivery?”

The team will also be examining the potential of harnessing these fully automated systems in city settings, for direct-to-customer deliveries.

Delivery robots are already starting to emerge in cities for short-distance local errands, often piloted by a remote operator who steers the vehicle with the aid of cameras. Residents of urban centers like downtown Los Angeles and San Francisco may have already stumbled across these friendly local bots wheeling about town delivering their goods. Swiss Post has similarly introduced small autonomous vehicles to deliver mail to communities in Switzerland.


A delivery robot in Redwood City, California. Image/Wikimedia Commons

“People have built the hardware. We have some version of this technology already, but as far as using them efficiently and the actual optimization of these systems — that’s very uncharted territory,” Carlsson said.

In Carlsson’s vision, these fleets of smaller “sidekick” robots, drones, or small vehicles could work with a delivery van, such as Amazon’s vans, to quickly offload goods and deliver them locally, allowing the van to continue its journey. He said that optimized symbiotic routing does not necessarily mean that the helper robot fleet is working for the one host vehicle.

“Maybe you have a robot that picks up a package from van number one, and then it drops it off, and then it goes to van number two,” Carlsson said.  “It makes a lot of sense to do things that way, because a robot is just moving along the sidewalks not cruising along a highway. Your robot may pick up a package from a van and drop it off, but by the time the robot is free again, that van is in another part of town.”

Carlsson said he believed there would be a future where symbiotic vehicles were an important part of everyday life, and his project’s goal in the first instance was to determine whether these systems could be helpful and workable.

“We’ve proven it from a theoretical perspective, but this is going to be much more simulation-driven and much more algorithmic,” Carlsson said.  “We want to have more than just the mathematical proof of this. We want to know; does it actually work?”

Published on June 22nd, 2022

Last updated on June 22nd, 2022

Share This Story