TwinCoBot
Industry of the Future is characterized by smart automation and connectivity, flexible and adaptive processes, as well as sustainable manufacturing practices.
To realize this ambitious vision, system models that aggregate individual digital models of components or subsystems (digital twins) and processes need to be developed to enable companies to meet the demands of the new industrial economic paradigm. The term “digital twin” is recent but the underlying concepts are not new. Digital twins have been shown to provide real benefits to the production life cycles of companies.
The goal of this project is to build digital twins of a critical subsystem composed of humans and cobots in a collaborative task for application in the smart factory of the future. In future, we can leverage the digital twins and the generic software architecture developed during this project to investigate cybersecurity, privacy, ethics in AI , health and wellbeing of workers, trust in automation, etc.
We will also be able to pursue major grant funding from ANR, European projects (Horizon 2022), etc. The team will be a in strong position to partner with companies to translate our technology for field deployment.
Expected outcomes
Deliverables of the project include:
- digital twins that include humans, with their physical and performance models, and cobots that are AI-enabled,
- software architecture to allow for learning (embedded AI) and adaptation to changes in the environment (dynamic variability). At the end of the project, we will have a set of common tools for working together (e.g., Robot Operating System ROS 2 or ROS Industrial), as well as the capacity to generate more quality data that can be applied to the development of future system models for Industry 4.0.
IMT Atlantique and partners Roles
We have assembled a multi-disciplinary team consisting of computer scientists, human factors engineers, roboticists, cognitive system engineers from three different research groups across the Carnot Institutes. The researchers bring complementary expertise and resources (Living Lab in Brest, HUT in Alès) to bear on the objectives of this project, while leveraging a common technical approach in software engineering and performance modelling. This unique combination allows us to accelerate our development work, by pooling data that are collected at the two sites.
This project is a collaboration between the INFO Departements of IMT Atlantique and the one from Mines Alès, with researchers from LABSTICC (P4S and RAMBO groups) and EUROMOVDHM (MIB). Research will be conducted in the LIVING Lab at IMT Atlantique and the HUman at home ProjecT (HUT) lab at Mines Alès. The two labs will replicate the selected task in two different environments, so as to create two similar datasets with different cobotic arms. Once the dataset are created, a software architecture for the two datasets will be developed. The generic architecture will enable connection of the different tools required for the creation of the digital twin, as well as for any artificial intelligence tools to be implemented. The digital twin generated from this project will be available for any smart factory or any industry that requires human-robot interaction