SWEET — Social aWareness for sErvicE roboTs

This project develops a security and privacy framework for social and care robots, especially in healthcare.

• •Privacy-preserving, user-centred robot interaction
• •Seamless and continuous authentication supported by adaptive biometric signals (e.g. gait patterns captured by the robot)
• •Efficient access control for unstructured data across multiple stakeholders
• •Scalability and reducing the risk of unintended data disclosure
• •Enhancing robots' understanding of cultural sensitivity and social behaviours to support trust and acceptance

This project advances Artificial Trust (AT) in Human-Robot Interaction by integrating Theory of Mind (ToM) into trust assessment models.

• •Combine observable behaviours, ToM insights, and internal trust factors into a comprehensive trust assessment framework
• •Enable robots to fine-tune responses based on human needs and preferences
• •Develop a novel artificial cognitive architecture informed by machine learning, psychology, and social sciences
• •Experimentally validate the approach with a humanoid social robot, assessing both trustworthiness estimation accuracy and its impact on task performance

This project investigates a software infrastructure to help robotics applications run across multiple robot types.

• •Enabling kinematic-compatible applications to operate seamlessly across platforms
• •Designing a shared memory, cloud-native, low-latency communication framework for remote robot access across research groups
• •Developing solutions for multi-entity robot control, supporting scenarios where multiple software systems share the same hardware resource