Robots in an assisted living environment

Simon O’Keeffe
Maynooth University

My research aims to develop systems that allow robots to operate in an assisted living environment. Robotics is a rapidly growing field. Every few weeks there’s a new robot released that is able to do things that no robot could do before that. Through my research I get to be part of that. I get to contribute to what new tasks robots are able to do and how well they can do it. That’s why I love my research.

Robots are really impressive, but they’re also extremely complicated. They are good at some things, like carrying out a specific task precisely or working in dangerous environments. However, robots struggle in dynamic environments where objects and people can move around. The tasks that robots are bad at are typically suited to humans, because humans can easily adapt to a changing environment to complete a task.

For example, consider a home robot tasked with going to the kitchen to fetch a packet of crisps. What does the robot do if someone is blocking the way to the kitchen? What if the robot can’t find the crisps? Or they have been moved? What if there are no crisps but there are biscuits? All these questions might seem simple to a human, but are difficult to a robot.

In my research, I try to help robots overcome difficult tasks by getting a human to help them. Through a smart human robot interface, a human could help the robot when it gets into trouble, gets stuck, or just isn’t doing something exactly right. I work on developing these interfaces and investigating different ways to combine human and robot capabilities for different tasks. My research uses Maynooth University’s robot soccer team as a test bed for a dynamic assisted living environment. Robot soccer requires robots to move around and manipulate objects in a dynamic environment, a lot of the same problems present within an active home.