Claire chatted to Samuele Vinanzi from Sheffield Hallam University about how robots can tell whether to trust or distrust people. Samuele Vinanzi is a Senior Lecturer in Robotics and Artificial Intelligence at Sheffield Hallam University. He specializes in Cognitive Robotics: an interdisciplinary field that integrates robotics, artificial intelligence, cognitive science, and psychology to create robots that perceive, reason, and interact like humans. His research focuses on enabling social collaboration between humans and robots, particularly emotional intelligence, intention reading, and artificial trust. His recent book, " In Robots We Trust ", explores trust relationships between humans and robots.

Trustworthy robots also increase their market success and would be perceived The acceptance of novel technologies and social robots has as being developed and used responsibly, mitigating many specifically been described as a critical factor to successfully of the ethical considerations and challenges outlined by the deploy socially interactive robots on a large scale [1], [2], interactive robotics stakeholders [23], [24], [25].

Humans are known to construct cognitive maps of their everyday surroundings using a variety of perceptual inputs. As such, when a human is asked for directions to a particular location, their wayfinding capability in converting this cognitive map into directional instructions is challenged. Owing to spatial anxiety, the language used in the spoken instructions can be vague and often unclear. To account for this unreliability in navigational guidance, we propose a novel Deep Reinforcement Learning (DRL) based trust-driven robot navigation algorithm that learns humans' trustworthiness to perform a language guided navigation task. Our approach seeks to answer the question as to whether a robot can trust a human's navigational guidance or not. To this end, we look at training a policy that learns to navigate towards a goal location using only trustworthy human guidance, driven by its own robot trust metric. We look at quantifying various affective features from language-based instructions and incorporate them into our policy's observation space in the form of a human trust metric. We utilize both these trust metrics into an optimal cognitive reasoning scheme that decides when and when not to trust the given guidance. Our results show that the learned policy can navigate the environment in an optimal, time-efficient manner as opposed to an explorative approach that performs the same task. We showcase the efficacy of our results both in simulation and a real-world environment.

If we are serious about long-term human presence in space, such as manned bases on the moon or Mars, we must figure out how to streamline human-robot interactions. Right now, even the most basic of robots seem to have impenetrable brains. When I bought an autonomous vacuum cleaner, one that roams the house on its own, I thought I was going to save time and be able to enjoy a book or a movie, or play longer with the kids. I ended up robot-proofing every room, making sure wires and cables are out of the way, closing doors, placing electronic signposts for the robot to follow and much more – often daily. I cannot fully understand or predict what the system will do, so I don't trust it.