Digital Transformation

Dive deep into the transformative influence of digitalization in Material Science and Engineering. This symposium, designed as a beacon for innovators and thinkers, will delve into every nuance of digitalization's impact, from pioneering predictive simulations to avant-garde materials design techniques. Together, we will look at case studies, share research, and forecast the future where nearly three-quarters of material-based innovations will be born out of digital advancements. Whether you're a researcher, industrialist, or academic, this symposium promises a comprehensive look at a digitally-empowered future in materials science.

While digitalization in Material Science and Engineering promises to open up new horizons in this research field, the challenges to introduce digital tools in everyday laboratory life have to be addressed alike. The introduction of electronic-lab-books for example do not only need a common understanding of how to structure and handle data according to the FAIR (findable, accessible, interoperable, reusable) principles, but also a user friendly design of the digital infrastructure. Moreover, one of the cornerstones of a successful digital transformation of laboratory life is to get the analysis and test equipment suppliers on board. And finally, educational measures for scientists and lab technicians have to be established. Best practice examples as well as introductory overviews of single solutions shall be presented in this symposium, bringing together the key enablers of everyday laboratory life, be they data stewards, or lab officers or providers of experimental equipment. 

Materials Science and Engineering is undergoing a major paradigm shift towards more efficient digitalization. Integration and reuse of data and knowledge from synthesis, production, characterization as well as of modelling activities open new perspectives for innovation. Emerging fields of Materials Informatics employing tools such as machine learning, big-data applications, statistical inference and Integrated Computational Materials Engineering (ICME) allow accelerating the discovery of new compositions and processes tailored to the production of materials with specific properties and microstructures. Efficient modeling and simulation of materials engineering processes is based on large amounts of heterogeneous experimental and simulation data. This data captures multiple scales, from atomistic to continuum, and a diversity of relevant physical, chemical and mechanical concepts such as thermodynamics, kinetics, functional and mechanical properties as well as metadata on materials history, data origin and provenance.

A key to enable the digitalization of materials and to leverage the advantages and opportunities of the digital age is an interoperable digital representation of materials and processes. An appropriate management of materials data requires the use of FAIR principles (findable, accessible, interoperable, and reusable). Digital workflows ensure the unity of materials data and used simulation protocols. They connect individual software tools, automatize the storage and curation of final simulation results as well as relevant intermediate steps and can, therewith, ensure the reproducibility of computational procedures. Ontologies are essential for formally representing universal materials science concepts, their interrelationships, and workflows. Application ontologies enhance identification, data integration and fully fledged complex simulation workflows. This will improve explainability and validation of real-life and simulated process designs. A unique identification and elucidation of entities and relations is required to meet the FAIR principles.

In this symposium, we call for an open discussion and exchange about the recent technical and scientific challenges involved in developing an interoperable representation of materials and processes. These include recent developments of ontologies, materials data schemas and software solutions that allow representation and integration of workflows, processes, and materials in a digitalized manner. Emphasis will be placed on the current developments towards a European Materials Data Space and its connections to similar national initiatives.

The acceleration of materials development tailored for additive manufacturing (AM) is of high importance in academia and industry to fully harness the capabilities of AM technologies and promote resource-efficient production. Data-driven design of materials and structures is emerging as a powerful new paradigm in this context. Advances in digitization as well as high-throughput experiments and simulations have significantly increased the availability of material data ­ laying the groundwork for computational methods to design, optimize, and inversely engineer advanced alloy systems. These approaches enable the targeted adjustment of composition and microstructure in both conventional materials as well as internal structures of architected materials to meet predefined performance criteria. Research in this area spans multiple scales and disciplines, implying, e.g. thermodynamic modeling for phase prediction, microstructure evolution simulations to capture solidification dynamics as well as techniques for image-based microstructure reconstruction. Computational homogenization of these microstructures further promotes the predication of the effective material properties. Beyond constitutive modeling, data analysis and machine learning play a pivotal role in extracting knowledge from simulations through surrogate models. These approaches are essential for predicting process-structure-property (PSP) relationships, thereby enabling the computational design and optimization of materials and structures. Moreover, the effective use of data critically depends on well-designed workflows and interoperable digital platforms, key elements of digital transformation that are not only enabling next-generation materials and process design but are already shaping the landscape of additive manufacturing.

The session will be particularly dedicated to the ongoing results from RTG 2868, however is open to contributions from other research activities in the field addressed.

Key topics of interest of this session include, but are not limited to:

• Data-driven design of metallic materials adapted to AM,
• Process simulations for AM to predict solidification, phase evolution, and microstructure development as well as resulting material properties,
• Microstructural characterization and reconstruction,
• Modeling and analysis of process–structure–property relationships,
• Deformation and fracture behavior of additively manufactured alloys and structures,
• Inverse, generative design methodologies and optimization of conventional and architected materials.

The fundamental relationships among processes, structures, and properties are critical to advancements in materials science and engineering, forming the basis for the development of innovative and high-performance materials. One particularly powerful analytical method for elucidating these relationships is symbolic regression. This approach enables the generation of mathematical equations that not only predict material behavior under various manufacturing conditions but also optimize vital performance characteristics. Unlike traditional “black-box” machine learning models, symbolic regression provides interpretable outcomes that yield valuable insights into the systems being studied. This method enhances our understanding of how different factors such as manufacturing conditions or chemical composition influence the final material state, all without relying on predefined models or assumptions thus contributing the acceleration of materials design. The upcoming symposium aims to foster comprehensive discussions on the application of symbolic regression in various sectors of materials science and engineering, encompassing both advanced numerical simulations and practical industrial applications. Contributions from academic researchers and industry professionals are highly encouraged, promoting collaboration and advancing the frontiers of material innovation.

The digital transformation is enabling the acceleration of the design, discovery, development and deployment of new functional and structural materials solutions. Traditional methods for developing new materials, such as the empirical trial-and-error method, cannot keep pace with the current development of materials science due to their long development cycles and high costs. Data driven approaches, especially machine learning methods, play an important role in materials science already today, and also in the immediate future. These include effectively work with high-dimensional data sets, prediction of material properties, high-throughput methods for determining phase diagrams and crystal structures, material design, and efficient and cost-effective methods for controlling material processes.

The current wide range of possibilities but also the limits of data-centric methods such as artificial intelligence, autonomous labs, high-throughput materials synthesis and characterization, and materials combinatorics will be brought together in this symposium.

Enabled by the exponential growth of data storage capacity and computational resources as well as the availability of open-source data analysis tools and Artificial Intelligence (AI), the fourth paradigm of science, namely data science, has taken off and impacts various disciplines including functional materials research. Large-scale combinational databases and impactful repositories are emerging, assisted by high-throughput synthesis/simulations and materials digitalization. Groundbreaking opportunities can be enabled by AI-assisted materials discovery, particularly promising for the highly disciplinary fields of functional materials research. It can guide the synthesis and discovery of new functional oxides and compounds with superior properties in fields of e.g. batteries, electrocatalysts, oxygen transport membrane and ferroic materials.

With the help of emergent data science and machine learning techniques, the composition -- processing -- microstructure -- property relationships of functional materials can be mapped out both forwardly and reversely, leading to forward inference and inverse design, respectively. For instance, the intrinsic physical properties can be statistically modelled using various types of descriptors derived from crystal structures, whereas the extrinsic properties can be statistically understood based on microstructure through Machine Learning surrogates. In particular, the inverse design of materials can be carried out based on high-throughput combinatorial screening, global optimization based on Bayesian optimization, and generative deep learning.

This symposium invites contributions from various domain expertise and on different length scales including synthesis, characterization, simulations and data analysis, who are in general data shareholders in the field functional materials research. It offers particularly a platform to exchange various application possibilities of data science tools and to showcase the best practice.

Artificial Intelligence (AI) is rapidly transforming the landscape of materials science and engineering, not only by accelerating the discovery and design of new materials, but also by reshaping the entire research process, from data generation to integration, analysis, and decision-making. This symposium will focus on the critical role of AI as an enabling technology within interoperable, automated, and data-centric research environments, emphasizing the development and adoption of FAIR data infrastructures, automation, and AI-integrated platforms for advanced materials research. A central goal of the symposium is to address the challenges and opportunities associated with the digital integration of heterogeneous data sources, from simulations to experiments, and their use in AI-driven pipelines. Contributions to the symposium will aim at demonstrating concrete implementations of automated workflows and state-of-the-art AI tools applied to relevant use cases, including autonomous experimentation. The symposium is proposed by a consolidated network of institutions from both academia and industry, actively collaborating through initiatives focused on digital transformation in materials science. In particular, the symposium builds on the experience of the COST Action "European Materials Informatics Network", a major initiative funded by the European Commission, bringing together hundreds of experts on the integration of AI, data technologies, and interoperable infrastructures for advanced materials. Gathering experts in materials informatics, AI, data infrastructures, and automation, this symposium will provide a platform for discussing technical innovations, best practices, and cross-sector applications for next-generation intelligent, data-driven materials research. The event will also foster collaboration between ongoing European and international efforts, contributing to the development of an open, interconnected, and AI-powered materials innovation ecosystem.

The digitalisation of materials research is driving a transition toward integrated, automated and intelligent workflows that accelerate the development of advanced materials. This symposium will focus on Materials Acceleration Platforms (MAPs), which leverage advances in laboratory automation, self-driving labs (SDLs), simulation, data management and machine learning to speed up the entire innovation cycle from material conception to device development.

We invite contributions showcasing innovative approaches to automating synthesis, characterisation and data analysis tasks, designing and orchestrating data-driven workflows, coupling simulations with experiments, and integrating machine learning into experimental planning. Particular emphasis will be placed on MAP implementations, strategies for handling heterogeneous data sources (including legacy systems) and tools that enhance reproducibility, interoperability and scalability in MAP settings. Further topics include the enhanced support for FAIR initiatives through automation in execution and documentation, the use of linked data and ontologies, as well as modular hard- and software architectures. We also encourage submissions exploring the use of large language models (LLMs) to facilitate more natural and flexible human-machine interaction within MAPs.

This session aims to unite researchers across diverse application domains from advanced functional materials to structural materials as well as engineers and computer scientists to share insights into the design and deployment of full-scale MAPs as well as individual automated modules. We especially welcome submissions highlighting best practices in platform development for fostering cross-disciplinary exchange. Contributions addressing both early-stage discovery and late-stage validation in practical settings are encouraged. By providing a vibrant forum, this symposium will advance the role of digital tools and automation in shaping the future of materials research.

This symposium aims to facilitate global collaboration on large-scale materials research data by employing harmonized terminologies and schemas aligned with the FAIR (Findable, Accessible, Interoperable, Reusable) principles for materials R&D data. 

Building on the successes of previous symposia held in 2022(KINTEX, Korea), 2023(Wiesbaden, Germany), 2024(Jeju, Korea), and 2025 (Tsukuba, Japan), the 5th symposium will return to Germany to strengthening understanding and engagement across Europe and worldwide. We will thoroughly investigate the current barriers to data sharing, considering not only terminologies but also optimized data structures for efficient, data-driven materials design and analytics. Furthermore, the symposium will feature extensive discussions on the practical aspects of data sharing across various sub-disciplines of materials science, ultimately driving forward global efforts in data sharing.