I’m honored to serve on the Program Committee of the IEEE BigData 2022 workshop on Digital Twins for Accelerated Discovery of Climate & Sustainability (ADoCS 2022).
Digital twins are digital representation of a physical products or processes where a real object exchange data with a virtual counterpart. Digital twins become possible due to the availability of real time stream of Internet of Things sensor data, large scale physical and engineering simulations and massive volume of data that enable AI model training. The volume and quality of data makes digital twins a “virtual digital sandbox” for testing scenarios, efficient AI model testing and accelerated insight in complex system behaviors. Specifically, AI models can be applied to digital twins both to learn, to control and to predict future behavior before a product or process is implemented.
Physics constrained Artificial Intelligence Networks and surrogate AI models can be applied in digital twins to ensure physical consistency, to enforce explainability and to speed up AI model training and deployment. Current initiative of digital twin spans from modeling airplane design to create digital replicas of the Earth, like the “Destination Earth” initiative. Most approaches are combinations of petabyte of data generated by Internet of Things, satellite data, weather/ climate models and/or physical simulations but currently they require extensive computations and data storage.
This workshop will discuss the role of big data architecture and processing, ideal connection of data with AI models, efficient simulations of complex and multiscale processes and the requirement for building, testing and validating digital twins.