Eric Krotkov’s workplace is the stuff of little kids’ dreams.

In the sleek, modern office space, futuristic robots scoot around the furniture, and multi-jointed metallic arms gingerly pick up and manipulate television remote controls and cereal boxes.

This is the Toyota Research Institute (TRI), a $1 billion research initiative launched two years ago by the world’s largest automaker to integrate artificial intelligence and robotics into people’s everyday lives in ways that, until now, have only been imagined.

As the research institute’s chief science officer, Krotkov oversees the pioneering effort. It’s the latest accomplishment in a career spent at the bleeding edge of robotics. But despite the thrilling work, Krotkov admits it wasn’t his first career choice. His dream as a kid was to be an astronaut.

“I wanted to be an astronaut from an early age, but to do that, I realized I had to be a pilot,” says Krotkov, speaking via video conference from his Cambridge, Mass. office. “And in order to be a pilot, you had to have good vision, and my vision has never been good. So from an early age, I realized, ‘I am never going to be able to do this. How can I get [to outer space] using a proxy or a surrogate?’”

He set his sights on outer-space robots, like the rovers navigating extra­terrestrial terrain that he doodled in grade school and like HAL 9000, the sentient computer in 2001: A Space Odyssey, a film Krotkov has watched multiple times.

His interest in the subject deepened in 1977 after enrolling at Haverford, with which he’d fallen in love when he visit­ed campus during a high school debate trip. Majoring in philosophy, Krotkov wrote his senior thesis on machine epis­temology, the idea that thinking itself is a form of computation. Since Haverford had no classes on artificial intelligence at the time, he took courses on the subject at the University of Pennsylvania. His studies there led to him pursue his Ph.D. in computer and information science from the university, where he focused on computer vision—the challenge of getting machines to interpret and under­stand images.

After earning his Ph.D. in 1987, Krotkov landed a research faculty posi­tion at Carnegie Mellon University’s famed Robotics Institute, the first robot­ics department at a U.S. university and one that has churned out many of the top luminaries in the field. Krotkov fit right in, helping the institute develop robots designed to infiltrate malfunc­tioning nuclear reactors and perform precise surgical procedures. Software he developed helped the Sojourner rover keep track of its position after it landed on Mars in 1997—helping Krotkov to realize his childhood dream of venturing to outer space. “I felt con­nected to Sojourner in a very real way,” says Krotkov. “It was like, ‘I am your grandfather.’”

Once Krotkov felt he’d maxed out his potential contributions to the field of planetary rovers, he took a job in 1997 with the Defense Advanced Research Projects Agency, or DARPA, the Department of Defense agency responsible for developing new mili­tary technologies. While he was only on staff at DARPA for three years, he would later continue to work with the agency as founder and president of Griffin Technologies, a consulting firm specializing in robotics and machine perception. He focused on developing tactical robots used in exploratory and life-saving procedures, including robots that defused roadside bombs in Iraq and Afghanistan. “To me, this was my single biggest professional accomplishment,” he says. “I have had men and women come up to me, buy me drinks, and say, ‘You saved my life.’”

Near the end of 2015, Krotkov was tapped to help spearhead Toyota’s bold new venture in autonomous cars and robotics. “It aligned with my ideals and my values,” says Krotkov of TRI. “We have the opportunity to save many more lives than I did with the tactical mobile robots at DARPA. The number of people who die or are injured in car crashes is astronomical, and I wanted to put a dent in that.”

Much of the attention and excite­ment around TRI has focused on how the organization is faring against Google, Tesla, Uber, and other major automak­ers in the race to perfect driverless cars. But the majority of Krotkov’s work at the institute has so far focused on a different but equally important goal: Developing in-home “support robots,” especially those designed to help seniors age in place.

“It is a gigantic problem for our world to have an aging society,” says Krotkov. “The Baby Boomers are about to retire. The parents in China who live in a one-child regime are soon retiring. It’s largely graying in Japan. It’s a huge problem, and there’s a huge opportu­nity for robots that can help people live in their homes rather than nursing homes, care facilities, or hospitals. If the robots we develop lead to a better quality of life for these people, that will allow Toyota to make a difference in many, many lives.”

So Krotkov continues his pioneer­ing work in robotics, this time not on machines that will venture into war zones or the far reaches of space, but that instead could soon be assisting people in homes all over the world. In truth, it’s a continuation of the trailblazing work he started so long ago.

“From A.I. to computer vision to robotics, it’s the very same road I started down as a junior at Haverford College,” he says. “Other people had midlife crises or have career changes. I feel like I have been going in the same direction the whole time.”