Date of Completion

2024

Document Type

Honors College Thesis

Department

Computer Science

Thesis Type

Honors College

First Advisor

Dr. Josh Bongard

Second Advisor

Amanda Bertschinger

Third Advisor

Lisa Dion

Keywords

Evolutionary Robotics, Evolutionary Computation, Robotics, Simulation, Computer Science, Anthropomorphism

Abstract

As our lives become increasingly intertwined with technology, it becomes imperative to study human-robot interaction (HRI) through an emotional lens. Humans relating to robots on an emotional level could lead to enhanced robot-human collaboration and broader acceptance of technology. In the future, robots that perform dangerous rescues, care for the elderly and act as therapists will all benefit from the ability to display interpretable emotional behaviors. The purpose of this project was to develop simulated robotic behaviors that are expressive and representative of certain human emotions using evolutionary robotics. The result is an emotive behavior evolved from a random behavior that people can then recognize as a specific emotion. The robot designed for this project was quadrupedal and doglike. It has a 47-neuron neural network with 16 sensor neurons, 15 motor neurons, and 16 hidden neurons. Through anthropomorphism, human subjects were able to view an evolved robotic behavior from the robot, and come to a significant consensus on what emotion the behavior might represent. These results indicate that people are capable of agreeing on emotional robotic behaviors. With the advancing integration of robots into society, the capacity of humans to universally interpret robotic behaviors becomes increasingly paramount in tandem. This research proves for the first time that it is possible to evolve a robotic behavior that people can collectively recognize as expressing a certain emotion.

Comments

The full contents of this thesis are available only in the Honors College office.

Creative Commons License

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 License.

Download

Share

COinS