Supervisor: Ilaria Tiddi (i.tiddi@vu.nl)
Description: Ownership of objects is a crucial concept in human society, influencing behavior, ethics, and social interactions. For household robots, understanding ownership is vital to prevent them from interacting with objects they should not touch or use. In this project, we aim to explore how robots can identify and distinguish objects based on ownership. The focus will be on how robots can store and retrieve identifying information about objects, such as visual features, location history, or contextual clues, to infer ownership and ensure appropriate interactions within shared environments like homes.
Task: This project will involve researching existing methods for object identification and classification in robotics, with a focus on their applicability to distinguishing ownership. We will investigate different types of identifying information, such as visual markers, object usage patterns, and spatial context, that could be stored by robots to differentiate objects. The project will also include developing or adapting algorithms to enable robots to infer ownership dynamically, based on factors such as user interactions and social cues. An experimental setup will be designed to test these algorithms in simulated or real household environments, assessing the robot’s ability to correctly identify and respect object ownership. A simulation environment will be provided for the testing and development purposes.
RQ: How can household robots store and use identifying information to distinguish objects based on ownership, and what methods enable them to infer ownership effectively in real-world scenarios?