32(ish) Questions with an MIT Robotics Researcher (and Actor!)
Alex, a fourth-year Ph.D. student in MIT AeroAstro researching human-robot collaboration, speaks with Rachel at MIT CSAIL about safe robot coworkers, task planning, and his acting work.
In this talk
- 1:09
Why robots struggle to work near people
Alex says close collaboration is hard because robots must avoid harming people and must respond to unpredictable behavior. One lab demo replans around moving obstacles a thousand times per second. He says better understanding of people is still needed before robots can work safely and fluidly around them.
- 3:50
Research and acting at MIT
Alex describes getting involved in campus theater, starting with an acting class he took on a dare. He says acting skills help with research communication, including presentations and talks. He also recalls playing Mark Antony in Julius Caesar.
- 6:16
Learning to share tasks with a robot
Alex’s first research project studied how humans and robots can learn who is better at different tasks and allocate work accordingly. In a painting task modeled on industrial assembly, a person and robot share five subtasks, with one task assigned flexibly. The system combines task planning with a diffusion policy for robot motion, and learns from repeated rounds how the person works.
- 10:09
Testing for real differences between people
Alex combines traditional scheduling methods with newer techniques such as diffusion, which he says can be useful when data or computing power is limited. His study involved 32 people and deliberately varied how they performed tasks, including which hand they used, to better reflect differences in real workplaces. He says whether the robot gets out of a person’s way affects how people feel about working together, alongside task speed.
- 11:39
A people-centered vision for robotics
Alex explains that his research spans MIT AeroAstro and CSAIL, which brings together researchers across departments. He wants robots to help people in real settings, rather than focusing only on capability. His career advice is to look beyond current robotics and AI hype and consider less obvious angles, such as human-robot interaction.
Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.