When asked “Why study robots?”, Antonio Loquercio, assistant professor in electrical and systems engineering, says that he himself is 10% robot. Diagnosed with diabetes when he was four years old, Loquercio has relied on a continuous blood glucose monitor to track his blood sugar levels and administer insulin throughout the day for most of his life. This machine that senses its environment and responds accordingly fits into Loquercio’s definition of a robot, a definition that he has been applying to his own research questions for the last six years. And while Loquercio wants to help design and build robots that can be used to solve real-world problems, such as removing barriers for people with chronic diseases, safely doing household chores and assisting in surgeries, he resonates even more with the potential robots have in answering fundamental questions on decision making and what it means to be intelligent.
Loquercio’s passion for robotics stems from his childhood curiosity and lifelong love of nature. Spending afternoons outdoors in Capranica, a small town outside of Rome, Italy, the young scientist would watch ants marching in line, birds building intricate nests and schooling fish moving as one.
“I was always fascinated with the way animals carry out specific tasks,” says Loquercio. “I wanted to know how and why they chose to perform those tasks over others and how they were picking up on cues from the environment to help make those decisions.”
“Studying the decision-making process in animals is very complex and there are many uncontrollable variables,” says Loquercio. “By building a separate robotic system, we can examine those questions at much higher granularity and study every level of decision making. This was what I started to explore in my Ph.D. through a very ambitious task: building the fastest drone possible.”
Teaching common sense to robots is no easy feat. At Penn Engineering, Loquercio is working to accomplish this goal through specific projects that aim to teach robots to walk, run and jump through different terrain while adapting to changes in their environment. He is also working on teaching robots to learn behaviors by imitating humans and teaching robots to make informed decisions with a limited set of prior information. But he won’t stop there. Once he accomplishes all that, Loquercio sees a clear next step in advancing robots to a point where they will truly be helpful in our everyday lives.
Read more at Penn Engineering Today.
