At the Farjam Research Group (FRG), we aim to understand, predict, and improve the behavior of complex dynamical systems, with a particular focus on advanced manufacturing processes such as nanomanufacturing. We view these systems as intricate, multi-scale networks where material behavior, process dynamics, and performance outcomes are tightly interlinked.

Understanding behavior starts with uncovering the relationships between process parameters, material properties, and final structure, commonly known as the process-structure-property relationships. We pursue this through carefully designed experiments, high-resolution characterization, and systematic analysis.

Predicting system behavior requires robust modeling frameworks. At FRG, we develop hybrid models that combine the strengths of physics-based and data-driven approaches. Physics-based models, grounded in first principles, offer interpretable insights without requiring extensive data but are often computationally expensive and may omit real-world complexities. In contrast, data-driven models offer scalability and adaptability but demand large datasets and can lack generalizability.

To bridge these gaps, we create physics-informed, data-augmented models that achieve high fidelity and predictive accuracy while maintaining computational efficiency and physical interpretability.

Yet even the best models cannot fully capture the variability and uncertainty of real-world systems. To address this, we focus on intelligent control strategies that adapt and respond to discrepancies between predicted and actual behavior. This includes the development of novel control frameworks, learning-based adaptation, and decision-making algorithms that enable real-time optimization and robust performance.

Together, our research advances the science of modeling and control for next-generation manufacturing systems pushing the boundaries of flexibility, precision, and reliability.

Semiconductors, those minuscule computer chips, power everything from our smartphones to our vehicles in our everyday lives. The escalating demand for semiconductors has become a pressing challenge for manufacturers striving to meet industry demands. The stringent requirements of semiconductor manufacturing, demanding precision at the nanoscale, unwavering reliability, and consistency, entail state-of-the-art semiconductor production involving intricate lithography steps like patterning, deposition, etching, cleaning, and doping within controlled cleanroom environments.

On the other hand, AM techniques offer cost savings and expand design possibilities across various materials, reducing complexity and minimizing material waste outside the confines of expensive cleanroom facilities. While AM for semiconductors empowers manufacturers to craft tailored components to suit their device/system needs, current AM processes largely operate open loop, leading to inconsistencies in behavior and falling short of semiconductor device standards.

At Farjam Research Group, we aim to address these core challenges by advancing scientific understanding, developing enabling technologies, and implementing intelligent control strategies towards the directed autonomy of these processes. This pursuit aims to harness and extend the realms of advanced manufacturing, artificial intelligence, and control theory, seeking to pivot the process of semiconductor manufacturing towards a more flexible and adaptive process that can be used to tailor the process parameters to meet specific semiconductor fabrication requirements around customized designs and non-traditional materials and surface interfaces.