January 31, 2023 9:15 - 13:00 (CET)
Center for Human Technologies - Istituto Italiano di Tecnologia - Fibonacci Room (Floor 10)
Join us on MS Teams
Abstract
The development of artificial intelligent systems as interactive instruments and collaborators of human well-being requires them to be designed on a human scale. To this aim, human awareness can be considered the system’s capability to interpret human behavior and be sensitive to their inner states, such as beliefs, emotions, or intentions. At the intersection of Cognitive Robotics, Machine Learning, and Human-Robot Interaction, this approach suggests that the robot should maintain a model of humans to understand, predict and anticipate their actions, needs and limitations. Informed of this, artificial systems could understand information more or less covertly expressed in human behavior. Also, this would allow robots to communicate and cooperate in ways that are intuitive for the human partner. From this perspective, human awareness is open to two other crucial features: explainability and trustworthiness. Human users need to understand how machines reach their decisions. This also concerns human-robot interaction, where explainable robotic behavior supports trust and safety from the human perspective. Mutual reliance and trustworthy interactions are allowed by mutual knowledge of the parties. Only by developing artificial systems with abilities to comprehend humans and to be comprehensible by them, the quest for a human-centered technology will be achieved. The workshop about this topic is promoted by the European twinning project TERAIS (Towards Excellent Robotics and Artificial Intelligence at a Slovak university), whose research aims to promote cognitive robotics and human-robot interaction through the following three pillars: human awareness, explainability, and trustworthiness.
