Shutterstock 2030685800
Shutterstock 2030685800
Shutterstock 2030685800
Shutterstock 2030685800
Shutterstock 2030685800

Data infusion enhances simulation software

Sept. 30, 2022
The availability of data and artificial intelligence is impacting simulation models and the digital twin. Prashant Srinivasan, director of AI products and applications at SymphonyAI Industrial, explains how to leverage that trend.

Prashant Srinivasan is director of AI products & applications at SymphonyAI Industrial.

Tell us about your company’s state-of-the-art simulation-software technology.

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial: SymphonyAI Industrial’s Performance 360 software features simulation capability using a combination of data-driven AI/ML models and first-principles-based physics models. The data-driven models leverage deep-learning algorithms such as recurrent neural networks (RNN), long short-term memory (LSTM), transformers and deep-state space models, while the physics-based models cover thermodynamic simulations, heat and mass balance equations, and equations of state. This software has been applied across industrial assets such as compressors, turbines, pumps, heat exchangers and furnaces, as well as for system-level simulations including refinery pre-heat train and combined cycle power plant.

What have been the biggest improvements to simulation-software technology in the past five years?

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial: The biggest improvement has been the infusion of data into simulation software in many different ways to enhance the outcomes—for example, the integration of real-time data with simulation models to improve model accuracy, as well as simulate closed-loop behavior and plant operations in a more realistic manner. A powerful application of this is high-fidelity 3D virtual-reality simulations for operator training that integrate real-time data.

Integration of big data from the cloud can calibrate and validate simulation models for a wide range of scenarios, leveraging powerful and scalable hardware on the cloud and run very exhaustive what-if-analysis scenarios in a reasonable amount of time.

What’s the most innovative or efficient simulation-software technology application you’ve ever seen or been involved with?

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial: We have developed an AI digital-twin model and a real-time optimizer for optimizing energy, throughput and quality in melting furnaces. The real-time simulation is based on explainable deep-learning models that can accurately predict or forecast temperatures at various locations on the furnace, taking into account all major inputs and disturbances in the process. The explainability feature of the model can give operators an idea of what variables are impacting the model prediction the most and why a certain action is being recommended by the optimizer. This is third-generation AI and brings in more context, insight and trust in the results of the simulation model.

How has simulation-software technology benefitted from remote connectivity and networking?

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial: Remote connectivity and networking have helped with seamless integration of data from manufacturing plants to simulation software via the cloud. It has also enabled remote collaborative development of highly complex simulation models accessing data from the cloud. Likewise, it has facilitated the smooth deployment of software updates/patches remotely. Finally, it has enabled running complex simulation models on the cloud, communicating data continuously with a manufacturing plant and providing rich insights and inferences to plant engineers and operators. Such solutions can easily be scaled to multiple similar plants across geographies, thanks to the power of networking and cloud.

Can you explain how improvements in simulation-software design and production have impacted industrial applications?

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial: Improvement in simulation-software design and production have certainly increased the adoption and usability of industrial applications. It has also improved the productivity of the users in general.

How do simulation-software technologies figure into digital-twin platform models being used by manufacturers?

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial (symphonyindustrial.ai): Simulation-software models can be integrated into digital-twin-platform models in multiple ways, either by running as a compiled object or plugin within the platform and interfacing with the model or by running standalone remotely and pushing outputs into a cloud storage, which are then accessed by the digital-twin model. Sometimes a high-fidelity simulation model from the software may be used to provide operational boundaries and training data to extract a reduced order model that can be deployed as a component of the digital-twin model on a platform.

When will simulation-software technology become IT-friendly enough that engineers/IT professionals are no longer required for installation and operation?

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial: The focus in the industry on more intuitive user interfaces and user-friendly self-service capabilities, as well as infusion of artificial intelligence into software technology, are some steps being taken in that direction. We expect that the transformation will happen over the next decade in a gradual manner. There will always be some niche technical areas where engineers or technical experts are still required to help users out.

What future innovations will impact the use of simulation-software technology in manufacturing operations?

Prashant Srinivasan, director of AI products & applications, SymphonyAI Industrial: The following future innovations could make an impact on simulation-software technology in manufacturing.

Integration of diverse types of data sources—unstructured text/logs, images, videos, voice—and artificial intelligence in simulation software will provide more context and insight into the results.

Innovations in user experience and human interface with simulation software will make the experience more intuitive and interactive for operators and users.

About the author: Mike Bacidore

Mike Bacidore is the editor in chief for Control Design magazine. He is an award-winning columnist, earning a Gold Regional Award and a Silver National Award from the American Society of Business Publication Editors. Email him at [email protected].

About the Author

Mike Bacidore | Editor in Chief

Mike Bacidore is chief editor of Control Design and has been an integral part of the Endeavor Business Media editorial team since 2007. Previously, he was editorial director at Hughes Communications and a portfolio manager of the human resources and labor law areas at Wolters Kluwer. Bacidore holds a BA from the University of Illinois and an MBA from Lake Forest Graduate School of Management. He is an award-winning columnist, earning multiple regional and national awards from the American Society of Business Publication Editors. He may be reached at [email protected]