Automating mankind: Penn State summer camp extends the human experience

UNIVERSITY PARK, Pa. -- Killing people is something that keeps Sean Brennan up at night.

“I have to live with the consequences of whatever choices I make right now,” he said.

Brennan is not a serial killer. He is, however, challenging students to think about the ethical quandaries of being responsible for human lives. Since 2003, he has been a professor of mechanical engineering in Penn State’s College of Engineering, but during June and July he leads a camp in the Science-U program, one of the Eberly College of Science’s largest youth offerings. The choices he’s making now involve how best to teach the next generation to think beyond the easy questions.

Every summer the college coordinates summer camps at the University Park campus for kids ranging from second grade to high school. Dedicated to fostering their scientific curiosity, the Science-U summer camps regularly have between 500 and 600 participants.

This year Brennan debuted a brand new camp, A.I. vs. Science-U, which focuses on teaching high school students the basics of robotics and programming and how to develop an understanding of robots as an extension of the human experience.

“The embedded intelligence within robots is starting to bring up some interesting ethical dilemmas,” explained Brennan.

The game changes when humans become the driving force behind the way computers “think.” Self-driving cars, once an imagined possibility for the future are now a reality. Cooperative adaptive cruise control, a system car manufacturers are implementing in test vehicles around the world, places control of the vehicle in the vehicle’s proverbial hands. Or rather, in the hands of the computer controlling the car.

“This cruise control system will set the speed to follow the car in front of you,” said Brennan. “That follow-the-leader will continue until either the human overrides the system, or the vehicle runs out of fuel.”

This doesn’t take all responsibility away from the driver, Brennan explained. Certain situations, such as a jackknifed fuel tanker or hazardous weather conditions, still require a human reaction. But this dual-control means car manufacturers could face lawsuits over who, or what, is at fault when a driver is injured or killed in an accident when the vehicle’s automated system doesn’t detect something dangerous ahead.

“Who’s at fault?” asked Brennan. “Is it the car’s automatic system or the person behind the wheel? Perhaps it’s the computer programmer who designed the system?”

There’s a reason Brennan is so instinctive about asking these difficult questions; he’s been doing it since he was a kid. The excitement of boldly going into a new frontier lured Brennan toward becoming an astronaut. Focusing his studies on math and science, he applied and was accepted into the U.S. Air Force Academy.

“I met a bunch of astronauts while I was there, and I realized something about them pretty quickly,” he said. “They’re very much instantaneous problem solvers. It’s the engineers on the ground who are thinking about the long term, big picture.”

And that is exactly where Brennan wanted to be, too. He became fascinated with how humans will eventually explore the far reaches of the universe.

“It’s not going to be with our fragile bodies,” he explained. “We’re going to build super-intelligent robots to withstand the hazardous conditions our bodies can’t.”

That exploration won’t happen without automation, which became one of Brennan’s areas of expertise. While attending a conference in 2003, a Penn State faculty member approached Brennan about a teaching position at the University. While here, a colleague asked him to bring some of his robots to a summer camp at Penn State Berks to demonstrate their automation capabilities. Brennan was a hit.

“I was only supposed to be there until noon,” he said, laughing. “By 4 o’clock they still couldn’t get the kids to leave.”

Requests for a few hours of demonstration turned into requests for a full day, then an entire week. Finally, Brennan was asked to coordinate and run his own camp during the Science-U summer camps at University Park, where he’s teaching others to think critically about the ethics of robotics.

One way Brennan plans on imparting these lessons is by giving his students an unsolvable problem.

“At some point, a car will be built that will make a decision of whether or not it has to save the passenger or kill the person outside the vehicle,” he said. “If you hit the pedestrian, they’ll almost certainly be killed. If you swerve to avoid them, the car could go over a cliff and kill the driver.”

Brennan challenges his students to think about these types of dilemmas. As future programmers, they’ll need to answer those unsolvable questions before they input decision-making algorithms into consumer vehicles.

Brennan said there’s one activity in particular that really gets students thinking. Camp staff will set up robotic vehicles to follow a loop in cruise control, changing the path of the lead vehicle to illustrate the capabilities of automated systems.

“There’s always one student who shoves their foot out in front of the lead car,” said Brennan. “And then all the vehicles behind it crash.”

Students are tasked to figure out who’s at fault, discovering quickly the question has no easy answer. Brennan said this is an opportune moment to discuss how poorly programmed vehicles can cause nearly instantaneous traffic jams, hindering the ability of rescue workers to reach those in need.

“There’s some liability for a programmer,” said Brennan.

Brennan conceded these complex ethical dilemmas can appear to be over high school students' heads, but explained he’s actually targeting them at the right time in their growth phases.

“We’re hitting these students right when they’re starting to recognize the need to be part of the community,” he explained. “A lot of the themes we talk about in this camp go back to the notion of helping others versus helping themselves.”

And these students will help others in the future, becoming the robotic engineers who program systems to explore inhospitable environments and guide consumer vehicles on self-driven road trips. Brennan is hopeful they’ll use the lessons from the safe environment of the A.I. vs. Science-U camp to save as many lives as possible in the process.

For more stories about IT at Penn State, visit Current at http://current.it.psu.edu.

Contacts: 
Last Updated July 18, 2014