New Call for Collaboration in Innovative Technologies

With technological advancements, companies now call for new collaboration in innovative technologies, largely focusing on digital twins and SciML.

FREMONT, CA: Today’s world encounters a call for collaboration in innovative technologies seeking to transform highly innovative, fundamental knowledge secured from advanced technological research into applied software technologies designed to substantially impact research across all disciplines. This call for proposals focuses on research from a domain-oriented research discipline with an emphasis on technology development or a technology-based research discipline such as data science, computer science, or AI, who have a profound interest in making their research results applicable to other disciplines. This call is open for proposals from primarily two technology areas—Digital Twins and SciML.

Digital Twins: Virtual Representations of the Real World

Digital twins, the virtual representations of real-world objects or systems, are based on a combination of several models, sometimes supported by machine learning and refined with real-time data. 

Accompanied by interactive analysis and visualisation, this technology opens up innovative research routes, enabling real-world modelling at an unprecedented scale and extreme detail level. Additionally, it allows researchers and policymakers to administer what-if scenarios, supporting decision-making. Some of the possible research topics include model coupling and integration, uncertainty quantification, improving modelling and simulation capabilities with machine learning and surrogate models, interactive analytics, and data assimilation.

SciML: Combining Machine Learning with Scientific Domain Knowledge

Amalgamating machine learning with domain knowledge yields models that operate with fewer data and are more efficient in computational processes and energy requirements. At the same time, they are more accurate and trustworthy. In addition, a method focused on capitalising on domain knowledge eliminates a significant traditional machine learning method limitation, which is learning only from what they see. Combining machine learning with domain knowledge will make AI more widely applicable and attractive to domain researchers.

Scientific ML (SciML) addresses domain-specific data challenges and acquires new insights from research data through innovative methodological approaches. SciML is not confined to the exact sciences but extends to applications in all research areas. It uses machine learning and scientific computing tools to develop new methods for learning and data analysis that are domain-aware, scalable, and interpretable.

Some of the possible research topics in SciML include non-traditional or low-cost data sources, data fusion, extracting insights from multiple sources and sensors, and Robust and reliable learning, which includes uncertainty quantification, stability, validation, performance metrics, and reproducibility. It also involves domain-aware and physics-informed learning, hybrid models that include data-driven and domain-aware components and interpretable or explainable machine learning.