Artificial intelligence is opening new possibilities in materials and process engineering. Understanding how a material is made (e.g., the synthesis process), what its internal structure looks like (e.g., its microstructure), and how it behaves in applications (e.g., its macroscopic properties) is essential for developing new advanced materials and technologies. Research and development of materials and engineering processes that purely rely on experiments and conventional simulations can become expensive in both time and resources, making it difficult to explore many design options or to find the optimal solutions.
AI methods can significantly speed up this process. Machine learning models can be trained to predict the performance of materials much faster than conventional simulations, and generative models (for example generative adversarial networks) can create realistic virtual 3D microstructures of materials based on relatively few experimentally measured examples, e.g., via microscopic imaging. This enables virtual materials testing and supports inverse design, where AI suggests how materials or manufacturing processes should be adjusted to achieve desired performance.
By connecting processes, structures and properties, AI can help researchers to design new materials and optimize production processes more efficiently: Accelerating innovation in energy storage, advanced manufacturing and beyond.