Digital Twin at Siemens
Digital Twin at Siemens
Digital Twin at Siemens
Amit Chakraborty, Ph.D.
Principal Scientist, Siemens Technology
Abstract:
A Digital Twin is a dynamic, virtual representation of a physical asset, process, or system that evolves continuously across its lifecycle by integrating real-time data, simulation, machine learning, and reasoning. Far beyond a static 3D model, it mirrors the structure, context, and behavior of its physical counterpart, enabling predictive insights and informed decision-making. Its scientific foundation combines sensor-based data acquisition, data integration, simulation modeling, machine learning for predictive accuracy, and reasoning for actionable intelligence, supported by rigorous Validation, Verification, and Uncertainty Quantification (VVUQ). The fidelity of a digital twin hinges on the quality of input data and the sophistication of its models, allowing it to forecast performance, detect anomalies, and optimize operations in real time. This creates a bidirectional feedback loop where changes in the physical system are reflected digitally, and digital insights enhance the physical system. Digital twins range from simple components to complex systems like factories, cities, or ecosystems, drawing on disciplines such as control theory, system dynamics, AI, data science, and computer graphics. Their applications span industries: manufacturing uses them to improve production and predict failures; aerospace for virtual design and maintenance; healthcare for personalized treatment simulations; urban planning for traffic, energy, and emergency management; and infrastructure for real-time monitoring of systems like power grids. By enabling predictive capabilities and continuous optimization, digital twins drive efficiency, reduce costs, and foster innovation, making them a transformative tool in the era of intelligent systems and smart environments.
Siemens is a major enabler for Digital Twin, thanks to its various simulation products that support multiple industries to create and maintain Digital Twins. In this brief talk, I will introduce some of the concepts, the range of Siemens products and finally touch on some advanced developments leading to Hybrid Digital Twins.
Biography:
Dr. Amit Chakraborty is a Principal Scientist with Siemens Technology. He has been with Siemens since 1996. Since 2003, he has been leading a Research Group, first as a Program Manager and since 2012 as a Research Group Head. Prior to that he was a Senior Research Scientist. As a Principal Scientist, Dr. Chakraborty focuses on the development of a technology portfolio in the emerging area of Hybrid Digital Twin that works at the interface of Simulation, AI/Machine Learning and control, with applications in Dynamical Systems, Condition-based maintenance/ Asset optimization, and Risk Modeling. While leading the Predictive Analytics research group, he and his team primarily focused on a variety of Industrial AI applications across several Siemens business units. He successfully led several large R&D initiatives for predictive maintenance and asset optimization of complex equipment (e.g., Gas Turbines and conveyor belts for automotive manufacturing), energy optimization for High Performance Buildings, Power Flow Optimization for Electric Grids and Semantic Modeling of customer and manufacturing data. In addition, he and his team led or participated in several DARPA and DOE proposals/programs.
Dr. Chakraborty received his M.S and Ph.D. in Electrical Engineering from Yale University in the US and his undergraduate from Indian Institute of Technology, Kharagpur. His research interests include Hybrid Digital Twin, Physics Informed Neural Networks, Dynamical systems, Reinforcement Learning, Optimization, and risk modeling. He has more than 80 US and International Patents and over 65 publications in respected journals and conferences. Dr. Chakraborty won the Siemens Lifetime Achievement Inventor of the Year award in 2021 and the Edison Award in 2022 and again in 2024.