Until this past summer, the closest Syed Saahir Ahmed, a senior at Penn State University, got to working with robots was through a student club. There, he designed simple machines to perform elementary tasks, such as navigating through a maze. But the aspiring electrical engineer had always hoped to work on more complex machines, and thanks to the U.S. Department of Energy’s (DOE) Science Undergraduate Laboratory Internship (SULI) program, he got his wish.
Ahmed works with three other college seniors — Anibal Morales, Matthew Krebs and Kevin Wandke, also a SULI intern — as a robotics intern at DOE’s Argonne National Laboratory. The experience gives him and his peers a deep dive into the field of robotics and its wide-ranging applications in manufacturing.
“Our goal is to create and demonstrate a robotic digital prototyping system that can enable more collaborative, more autonomous and more intelligent manufacturing.” — Young Soo Park, Argonne mechanical engineer
“The things we’re working on now are the things we’ll see being implemented in manufacturing within the next 5 or 10 years, so in many ways we’re ahead of the curve,” said Wandke, who is from the University of Illinois at Urbana-Champaign and also a SULI intern. “What better place to prepare for the future?”
Wandke, Ahmed and their peers are each integrating new technologies into a teleautonomyrobotic system, including augmented reality, machine learning, 3-D simulation and pliant materials. Their work builds on Argonne’s recent innovations in robotics and remote systems to safely operate, repair and design nuclear facilities.
Led by Argonne mechanical engineer Young Soo Park, the group is adapting robotic remote systems technologies designed for harsh nuclear reactor plant conditions, to enable deployment in manufacturing facilities that produce materials for use in harsh operating environments. The effort is part of a new manufacturing science and engineering initiative at Argonne that enables students to work on advanced technologies to improve manufacturing process efficiencies and develop new optimized materials for products.
“Our goal is to create and demonstrate a robotic digital prototyping system that can enable more collaborative, more autonomous and more intelligent manufacturing,” Park said. “We work to achieve this goal by focusing on integrating emerging technologies like augmented and virtual reality, simulation, artificial intelligence and advanced sensing, through an agile network.”
This research is expected to deliver transformative technology and processes to improve the efficiency and competitiveness of U.S. manufacturing, which is a key mission of DOE’s Office of Advanced Manufacturing. Argonne has paired internship students with R&D projects in the advanced manufacturing internship program as part of its growing manufacturing program.
Virtual modeling and simulation
In manufacturing, computer simulation can be a powerful and cost-effective tool for testing and optimizing facilities, robotic hardware and manufacturing processes ahead of production. Ahmed, who studies electrical engineering, tackles this challenge by building a model of a virtual nuclear waste-handling facility that can complete several robotic tasks that require fine precision, such as cutting and welding. Ahmed works to refine these actions, to make them more realistic within the simulation. He will also use the simulation to demonstrate and control these actions on Argonne’s robot Baxter.
“What I enjoy most is learning about the software used for the simulation; it’s easy to see how it can be applied to modeling many other facilities, not just nuclear power plants or packaging facilities,” Ahmed said. “I can adapt this knowledge to work in many different environments, in manufacturing and elsewhere.”
Manufacturing traditionally uses hard, metal-based robots, which can pose risks to humans and are not as good at performing delicate tasks. Soft robotics, a field that employs pliant materials, such as silicone, has the potential to overcome these shortcomings to improve the safety and efficiency of manufacturing processes.
Among the biggest advantages of soft robots is their ability to make it easier for humans and robots to occupy shared spaces without as much risk. Wandke, who is studying mechanical engineering, is exploring potential applications by prototyping a soft robotic hand for Baxter, a two-arm robot used for demonstration, and developing the human-robot interface to control it.
“The thing I really enjoy about working on this project is that it brings together diverse skill sets. In school, I might be working on problems with other mechanical engineering students, whereas here I’m working with my peers in computer science and biomedical engineering for example,” Wandke said.
“The diversity is important because a lot of the problems we’re looking at are so complex, you need to tap into other areas of expertise to come up with a solution, and Argonne is one of the best places to do that.”
Augmented reality and artificial intelligence
In manufacturing, improving collaboration between machines and humans requires better perception. What a human or robot cannot “see” or reconstruct, they cannot control or account for. Augmented reality — technology that blends virtual components with real-world perception — offers a novel way to improve humans’ ability to perceive and control the machines they operate.
Computer science student Matthew Krebs, a senior at Lewis University, is learning firsthand how to integrate and optimize what we see in augmented reality with what exists in the real world. He does so with the help of machine learning and 3-D sensor technology.
“There are so many ways we can apply what we’re learning here to make manufacturing faster and smarter, safer and less costly. That’s why having the opportunity to do things like machine learning — which I’ve never done before — has been extremely valuable,” Krebs said. “This has been the best experience to help me make an educated decision about my future.”
Augmented reality and 3-D sensing
Along with better perception, proper alignment of human action and robot action is vital to improving collaboration between humans and machines. Humans operating machines through augmented reality must be confident that their motions are aligned with the robot’s actions.
Morales, a senior at Florida International University who is studying biomedical and electrical engineering, is evaluating how well augmented reality and real-world actions align with each other, using 3-D sensors and a tactile device that controls the two-arm robot Baxter. He evaluates the robot’s performance under various test conditions to see if its arms move in time and in position with the human controller’s and looks for ways to improve coordination of motion.
“In school, the focus tends to be on the need to graduate, whereas here people really care about making new and useful technologies that can help the nation, and they encourage you to explore and test new ways to do so,” Morales said. “It feels good to be here because of that.”
The SULI program is supported by DOE’s Office of Science. The robotic work on nuclear facilities is funded by the DOE’s Environmental Management Technology Development Office. Leveraging existing research to enhance advanced manufacturing — the development of more efficient manufacturing processes or materials — is one of Argonne’s key missions and the thrust of the Manufacturing Science and Engineering initiative.
Source : Argonne National Laboratory