Securing Our Virtual and Real Worlds

March 26, 2019
Hertz Staff
Livermore, Calif

For computer scientist Kathleen Fisher (’94), the best security is the sort that can be guaranteed, mathematically. With cars, elections, infrastructure, and military drones increasingly computerized and networked, knowing with a high degree of confidence that hackers can’t get into a system becomes a matter of life and death.

Fisher studies how to use “formal methods,” which are a way of proving or disproving the properties of computer systems mathematically. It’s as if “a machine can look over your shoulder and make sure that you aren’t making any mistakes,” says Fisher. “That’s really important when we think about cybersecurity,” she says, because humans designing software are prone to making just the sort of mistakes that hackers can exploit.

Fisher demonstrated the usefulness of formal methods spectacularly as a program manager at DARPA, where she led the development of what one of her colleagues called “the most secure UAV [unmanned aerial vehicle] on the planet” by building it from the ground up with code that could be proven secure. The same techniques could be adapted to help secure driverless cars, election systems, and energy grids, all of which are vulnerable to hacking.

Now Chair of the Computer Science Department at Tufts University, Fisher continues researching applications of formal methods, including new programming languages for handling large amounts of incoming data. She also leads efforts at Tufts, and within the wider computer science community, to foster diversity in computer science, by sponsoring programs to mentor women in computer science, and redesigning the introductory computer science course at Tufts to be more fully accessible to students with no background in computer science.

“Women get the message that they’re not good at or not welcome in computer science,” she says. “We want to send the message to women and underrepresented minorities that they belong in the computer science community.”