Robotics Expert Foreshadows the Future at Polykarp Kusch Lecture

Robots are becoming ubiquitous. Food delivery robots roll casually throughout The University of Texas at Dallas campus. Lawn-mowing robots crisscross the field outside the Administration Building. Drone racing is now a team sport with its own student organization. What’s next?

At the 2023 Polykarp Kusch Lecture, Dr. Mark W. Spong, professor of systems engineering in the Erik Jonsson School of Engineering and Computer Science and holder of the Excellence in Education Chair at UT Dallas, explained the history of robotics as well as his prediction for the future to a University-wide audience that included President Richard C. Benson, Provost Inga H. Musselman and other leaders as well as faculty and students.

“My work has been at the intersection of control theory and robotics,” Spong began, as he described the breathtaking speed of technological advancement throughout his career. “Typically, I would speak to more technical audiences. Today, I’m going to spare you all the mathematics and kind of give you a higher-level presentation. I’m going to talk a little bit about the past — how you define a robot. Then for robotics today, I want you to pay attention to something called Moravec’s paradox and autonomy. Finally, I’m going to give some predictions and look at human-robot interaction.”

Since 1985, top faculty at UT Dallas have delivered the Polykarp Kusch Lecture. The lecture series is named for Polykarp Kusch, a Nobel laureate and professor of physics, with the intent to inspire the lively mind. Kusch was a celebrated teacher and a prominent physicist who left Columbia University to work at UT Dallas through its early years as a state institution and was a professor emeritus at the time of his death in 1993.

Dr. Stephanie G. Adams, dean of the Jonsson School and holder of the Lars Magnus Ericsson Chair, introduced Spong as the author of a textbook on robot modeling and control that has been used broadly by students internationally for more than 30 years.

“His solutions have stood the test of time and became the new foundation and standard in the field,” said Adams, also a professor of systems engineering. She described Spong, who served as Jonsson School dean from 2008 until 2017, as an educator at heart who devised robots that could play chess and air hockey and that would be more appealing to his audience of students. Spong also founded UTDesign®, a signature senior design capstone program that successfully matches companies with student design teams.

Spong has numerous accolades in engineering including fellow of the International Federation of Automatic Control and the Institute for Electrical and Electronics Engineers (IEEE). He has received the IEEE’s Third Millennium Medal, the Nyquist Lecture Prize and the Rufus Oldenburger Medal from the American Society of Mechanical Engineers and many more.

Robots Throughout History

Robots are now seen everywhere from manufacturing to medicine to household items including a “robot cat litter box,” Spong quipped. But how did they get their start?

Spong provided a brief tour through ancient history and popular culture to answer the question — what is a robot? The concept of a mechanical humanoid that could perform tasks for people had been part of human culture since as early as ancient Egypt, Spong said. Notably, ancient Jews had created golems, or animated anthropomorphic creatures made of clay or mud.

Later, during the Industrial Revolution, Mary Shelley’s “Frankenstein and Johann Wolfgang von Goethe’s The Sorcerer’s Apprentice” revealed that inventors who sought to recreate the human form were viewed with suspicion. More recently, Spong said, the name “robot” is traced to a 1921 play by Karel Čapek entitled “Rossum’s Universal Robots”, which likely was influenced by the original concept of the golem.

In the play, the humanoid robots were initially content to perform their tasks, but eventually, they rebelled and annihilated the human race. Though the play was produced more than 100 years ago, the public today may have familiar fears about new technology.

But how did inventors reach the modern definition of a robot?

Spong explained first that the commonly used definition of a robot is similar to the definition of a feedback control system.

“What is a robot?” Spong asked. “A machine that transforms sensing into action through feedback control capable of manipulation or locomotion.”

From a die casting machine used at a General Motors plant in the 1960s to today’s proliferation of robotic technology for household use, robots are built and controlled using similar general principles.

Robot Control and Technology Today

Today’s robots are substantially more sophisticated than they were just 50 years ago. Robots are able to accomplish well-structured tasks in well-structured environments, according to Spong. They can sense people, cars and other obstacles and exert simple control.

“There have been great strides made in walking robots,” Spong said, “no pun intended.”

Ironically, Spong explained that robots are better positioned to execute higher-order tasks such as accounting, editing or even generating code. These tasks are challenging and time-consuming for humans. On the other hand, sensorimotor tasks, particularly low-level manipulative tasks like folding a pile of laundry, are easy for humans but substantially more difficult for robots to perform efficiently. This phenomenon is known as Moravec’s paradox.

“Skills that appear most effortless to us are actually the most difficult to reverse engineer,” Spong said.

Spong’s primary focus of study has been robotics and control theory, specifically systems that allow robots to respond to external stimuli. In his Laboratory for Autonomous Robotics and Systems, students develop distributed control systems for drones and other robots.

“There’s nothing artificially intelligent about it,” Spong said after showing a video of one of his PhD students’ thesis presentations, demonstrating a robot playing air hockey. “It’s just a basic robotic control system taking in sensory data and an algorithm to program the path. But we included a confidence measure based on our sensory data. The behavior was hard to predict.”

Robots of the Future

In the future, Spong believes scientists will focus more closely on humanoids, networks and human-robot interaction. People will likely choose robots that resemble current robotic vacuums but still interact with them like humans. Spong shared an example of a family that had named their robotic vacuum and insisted on repairing the particular robot instead of replacing it when it broke, showing the emotional connection people can develop with robots.

Spong also presented an example of an interactive robot that could stand in for a person. Instead of moving a meeting to an all-digital format, a robot avatar with a video screen could allow a remote participant to move down a hallway and engage in water cooler chats with others in the office.

“Rather than replacing humans, I think the next step is for robots to augment human productivity,” Spong said.

Additionally, artificial intelligence (AI) and human interaction will continue to play a role in shaping the robots of the future.

“AI doesn’t create the robots themselves, but it can improve their performance in well-structured tasks,” Spong said.

An Ongoing Conversation

Faculty and students from across the campus attended the lecture, and they asked questions about ethics and sustainability.

Willie Chalmers III, a computer science senior, asked, “With increasing automation tools like ChatGPT and Dali, I’m curious what role traditional robotics will play.”

Spong answered, “Well, certainly the nature of research is changing. My work is physics-based modeling and control. The AI system learns through repeated trials. I see more incorporation of AI into the planning, control and perception aspects of robotics.”

In the future, inventors will remain focused on creating useful, animated helpers that can improve the human condition while considering disruptions that will occur with such dramatic changes in technology.