The ethics and social consequences of AI & caring robots. Learning trust, empathy and accountability (2020 - now)
Funded by the Marianne and Marcus Wallenberg Foundation under the WASP-HS - 2020-2024
As social robots will provide us with health and social care at different stages of our lives in the future, they must be abe to build trusting relationship with people and act and be designed in a manner that is ethically acceptable. The project focuses on understanding how trusting relationship are developed in three scenarios: educational robotics in schools, robots giving job interviews, and care robots for elderly people. Within this project, I am primarily focusing on analyzing and formalizing the process of attributing gender to social robots both from a design and a user perspective.
COIN: Co-adaptive Human-Robot Interactive Systems (2018 - now)
Funded by the Swedish Foundation for Strategic Research (SSF) - 2016-2021
COIN is a project that aims at developing a systematic, bidirectional short and long-term adaptive framework to yield safe, effective, efficient, and socially acceptable robot behaviors and human-robot interaction. Within this project, my task is to develop novel paradigms of human-robot affective co-adaptation (i.e., the bidirectional process of adaptation of a robot to the states of humans) to apply to educational scenarios. We are currently implementing the model of engagement I developed for my PhD to a geography literacy map game.
Modelling Trust in Child-Robot Educational Interactions (2018 - now)
As part of Natalia Calvo-Barajas' doctoral project within the MSCA-ITN ANIMATAS (H2020)
This doctoral project explores the effect of robot’s behavior on children perception of trust in educational settings. So far, we have investigated whether a robot’s facial expressions of happiness and anger could affect children’s perception of trustworthiness and how children perceive a robot’s trustworthiness when playing a storytelling game with it.
Persistence of the Uncanny Valley (2018 - now)
As part of Maike Paetzel's doctoral project
As the extant work on the uncanny valley mostly focuses on the influence of first impressions (based on the robot’s appearance) on the perception of uncanniness, we want to understand whether longer interactions with robots can help people overcome initial uncanny feelings. We are doing so both by increasingly exposing people to a robot’s interactive abilities and by manipulating the robot’s interaction strategies in terms of its behavioral patterns.
Facial Mimicry of Virtual, Human, and Robotic Agents (2018 - now)
In collaboration with Institut des Systèmes Intelligents et de Robotique, Université Pierre et Marie Curie
For this project, we are studying how people mimic the facial expressions displayed by human and artificial agents (virtual and robotic) and whether their mimicry changes based on embodiment and humanlikeness of the agent. In addition, we are trying to understand whether certain personality and empathy traits are predictors of people’s ability to unconsciously mimic a robot.
ENGAGE-DEM: A Model of Engagement of People with Dementia (2014 - 2018)
Funded by the Erasmus Mundus Joint Doctorate (EMJD) programme in Interactive and Cognitive Enviroments (ICE) - FPA n. 2010-2012
Within this project, I developed a model of engagement for people with dementia based on non-verbal behavior and psychophysiology. Such a model can be used to measure the inherent engagement of people with dementia during activities (e.g., interactions with social robots) in an exhaustive way and to assess the progression of engagement over the duration of an activity. The model’s overarching ambition is to lay the foundations for the automatic recognition of engagement in people with dementia. The achievement of such a milestone could bring to the design of robotic and interactive technologies able to recognize the engagement state of the person with dementia online and adapt accordingly.
Social Robots Supporting Narratives in People with Dementia (2012 - 2013)
In collaboration with the Robotics and Learning Technologies Lab of the University of Siena
For my Master’s thesis, I devised a semi-experimental study aimed at evaluating whether a social robot, the arctic seal PARO, could have a positive influence on the narrative capability of elders affected with dementia. The study followed a repeated measures design, and participants were asked to create stories in two conditions, with the help of static images (i.e., pictures) and with the robot. The sessions with the social robot showed an increase in positive affect (i.e., decrease in the occurrence of challenging behaviors and increase in happiness) and social interaction (i.e., higher number of questions and interactions with others). Moreover, the stories created with the robot were twice as long as those created with the pictures, and more autobiographical.