Building A Better DeliveryHelper

Marco Morales talks about his independent project in Northwestern Engineering's Master of Science in Robotics (MSR) program and how he's applying lessons from it to his internship work at Argonne National Laboratory

Marco Morales demonstrates his independent project.

 

Marco Morales was walking back from class in Northwestern Engineering's Master of Science in Robotics (MSR) program one day when he saw an Amazon driver making multiple trips back and forth from a delivery truck, the driver's arms laden with awkward, bulky packages. 

Morales (MSR '22) watched and wondered how, if at all, the driver's job could be made easier.

Morales used his MSR independent project to pursue an answer. His solution was a robot he dubbed DeliveryHelper. During the design and development process, he realized his creation could help more than just delivery drivers. 

“DeliveryHelper can help those who are elderly, have difficulty carrying multiple packages, or cannot walk a long distance while carrying items all day,” he said. “It can be used as a tool for assistance, not job replacement.” 

Morales’s creation came from a desire to build something that used multiple systems functioning together. The project required manipulation, navigation, and perception. DeliveryHelper is built on a Ridgeback mobile base with a Sawyer arm and Bumblebee camera mounted on it. The robot is designed to follow a person wearing an AprilTag on their back. 

Having each piece of DeliveryHelper do its specific job wasn't the hardest part for Morales. Getting the different systems to work together wound up being the biggest challenge. 

“I spent the most time trying to figure out how the robots communicate with each other,” he said. “One of the greatest things I learned from MSR is how to go about debugging errors. Sometimes the errors I received were simple and easy to fix, but others were vague and it would turn into a game of finding where things were failing.”

As DeliveryHelper progressed, so did Morales. He found immediate ways to apply what he was learning in his independent project to his work as a robotics research assistant with Argonne National Laboratory, where he’s been interning since May 2021. 

Argonne is a science and engineering research center created in the 1940s to help with the Manhattan Project that now seeks answers to some of humanity’s most challenging questions. Morales’ project there is to develop an even more advanced robot than DeliveryHelper that can navigate through a hazardous environment to retrieve an isotope. 

“I am actually applying a lot of the knowledge I gained on this project for my project at Argonne,” he said. “It involves many of the same aspects such as manipulation, navigation, and computer vision, but with more moving parts.” 

In an ideal scenario, Morales would love to make additional enhancements to DeliveryHelper. He'd like to have the robot not rely on detecting a large marker on the back of the person it was helping, and he also would want to incorporate computer vision to help differentiate packages based on size and shape. 

For now, though, he's happy with how the project turned out, and he's excited to continue applying the lessons he learned during the rest of his time in MSR — and beyond.  

“I think the project went well,” he said. “I was able to meet my goal, but I think the experience I gained was the most valuable aspect.”

McCormick News Article