Human Digital Twins in Hybrid Intelligence Systems (HD-THIS)

Abstract

The aim of the workshop is to collect and discuss research ideas about methods and tools to support Hybrid Intelligence with Human Digital Twins (HDTs). Hybrid Intelligence is defined as the combination of human- and artificial intelligence (Akata et al., 2020). It relates to the interaction or collaboration of humans with AI systems. As such and because it requires knowledge exchanges with common understanding between humans and AI systems, it should be grounded by research in human modelling and Knowledge Engineering. Hybrid Intelligence systems can assist humans in a Cyber-Physical-Social-System (CPSS) environment to accomplish tasks more efficiently and contribute to tackle a key challenge of today’s economy, which is re- and upskilling of the workforce. As digital representations of our world are created by the industry as Digital Twins, they focus on virtual representations of engineered systems (e.g. factories, cities, buildings, machines) and tend to forget the essential element: human-beings (in the role of workers, citizens, inhabitants or users). But quite recently has appeared the concept of the Human Digital Twin (HDT) (Naudet et al., 2021) which emulate characteristics and behaviors of humans, synchronizing their states through sensing and feedback loops. By providing a personal digital representation of a human-being, HDTs allow direct interaction with AI systems in the digital space and can be a key technology to improve interaction and collaboration within CPSS environments. This opens the door to Hybrid Intelligence systems, where humans and artificial intelligences interact or collaborate towards a common objective. Interesting research questions arise, such as: how HDT can be created; how HDT can support human-AI system understanding; how the latter can adapt its behavior to the human, and how to augment human capacities on a higher level than before.

Goals

The workshop aims at bringing together researchers and practitioners interested in advancing and implementing Human Digital Twins to enable Hybrid Intelligent systems. We expect an interdisciplinary audience that includes researchers from different backgrounds like cyber-physical systems and industry 4.0, digital human modelling and ergonomics, user modelling and personalisation, artificial intelligence, (vocational) pedagogy, psychology, and multiple fields of expertise like e-learning, knowledge management, human resources, among others. Authors of paper contributions will have 20 minutes to present and discuss their work. The workshop also includes an interactive part with the aim to collect ideas and topics from the audience to build a research agenda for HDT and Hybrid Intelligence in a World Café methodology (parallel, rotating groups). The workshop session will be open to registered participants of the PETRA conference.

List of Topics

In the scope of this workshop, we are interested in the exploration of the concepts, application scenarios, and development of tools. While our focus at LIST lies on personalised health recommendations and upskilling, we welcome perspectives from other domains to jointly develop a research roadmap.

Possible topics for contributions are:

• Models and theories for the Human Digital Twin
• AI methods for Digital Human Modelling
• Applications of Human Digital Twin for personalised assistance and learning
• Digital Human Modelling and its role in HDT implementation
• Human Digital Twins in a Metaverse (including MR, AR, VR)
• Collective Intelligence systems based on human models and ontologies
• Intelligent assistive environments involving human models
• Human-Robot Interaction and the role of HDT
• Human-AI collaboration
• HDT-supported learning and upskilling in Industry 4.0

Workshop Organizers