A social scientist by training, Leila Takayama stumbled upon robotics. The computer science department at UC Berkeley was building interaction design prototyping tools and Takayama’s job as an undergraduate was to run experiments that tested whether the tools were easy to use and whether they helped people be more open in their feedback.

As a social scientist, she was trained to do those type of experiments. She helped the team understand how to get honest answers from people by observing their behavior—not just what they say—and how different versions of their tools affect people.

This overlap between robotics and the social sciences is ripe with opportunities. As Takayama explains, “There are tons of theories in human psychology, in things like visual perception and human memory, that should increasingly inform the design of user interfaces.” 

In her work, Takayama acts as a bridge between the two.

“Psychology doesn’t always get applied to the design of technology,” she says. “I see myself as trying to translate what we know about humans and what that means about how we can design computer interfaces better.”

Now an Acting Associate Professor at UCSC, Takayama was recently awarded a Faculty Research Award from Google to study what it takes for a robot to learn from its human peers. 

For instance, if a delivery robot gets stuck on a sidewalk, what if it could get help from bystanders? Takayama and her team want to better understand how to design robots in such a way to make us feel some empathy for them—just as we would someone who was lost downtown.

Another example of the robot/human interaction that fascinates Takayama is Relay, a hotel robotics system that delivers things from the front desk to a guest room. After each delivery, a small screen asks guests how their stay is going. 

“People give it feedback and the feedback is helping the robotic system to learn what is good,” Takayama says. “Was it a good delivery? Was it a bad delivery? If it was bad, what was bad about it? How does it improve its behavior over time to make people happy? How do you provide good service to people?”

The robot can learn on its own with a little bit of training data, she explains, but sometimes it needs to check in with people to see if it’s doing the right thing.

At the heart of Takayama’s work is “getting more humans in the loop who can give useful feedback to the systems so they can learn to behave better.”

For instance, if a food delivery robot was rude, barreling down the middle of a sidewalk, someone could tell it, “Hey, that’s not cool, do it better next time.” This sort of interaction is helpful to the system developers.

“If you’re going to deploy a robot onto a sidewalk in a busy city,” Takayama explains, “you need to follow the norms of that city—and its different from place to place.” 

She wants to help robots and their designers better understand how to do that well. Her hope is that by being in that design cycle with designers and roboticists, social scientists like herself can help make robots a little more appropriate for the situations they’re being dropped into.

The big picture vision for Takayama is to bridge robotics with everyday life.

“Robotics is becoming more viable in the real world, but it’s still kind of researchy,” she says. “I hope that my work will help it be more human-centered—actually designed for normal people, not just roboticists.”

As for robotics and tech in Santa Cruz, Takayama would like to see more connection and collaboration between UCSC, startups, big companies and people building tech.

“I think we could help,” she says. “We’re teaching students how to do user experience research and interaction design. I’d be excited to see those skills put to use in these companies that are building real things. I want the research we do to get out into the real world.”

Are you a tech company in Santa Cruz? Become a Santa Cruz Works member at santacruzworks.org/membership

Cat Johnson is a storyteller and content strategist focused on coworking, community and living our best lives.