The Kitchin Group, based in the Department of Chemical Engineering at Carnegie Mellon University (CMU), is a leading research entity focused on leveraging data science, machine learning, and artificial intelligence to address challenges in catalysis and chemical engineering. Led by Professor John R. Kitchin, the group has made significant contributions to computational methods in materials science and is at the forefront of emerging paradigms such as agentic science. This report provides an overview of the group's structure, research activities, and notable achievements, with particular emphasis on their work in agentic science—a transformative approach to scientific discovery involving autonomous AI agents.
The Kitchin Research Group operates within the Chemical Engineering Department at Carnegie Mellon University, a renowned institution for engineering and technology innovation. The group is directed by Professor John R. Kitchin, who holds a PhD and has been a faculty member at CMU since 2006. Their primary location is Doherty Hall A207F, Pittsburgh, PA 15213, with contact available via phone at 412-268-7803.
The group's mission centers on utilizing data science and machine learning to solve complex problems in catalysis and engineering. This interdisciplinary approach integrates computational tools with experimental design, fostering advancements in sustainable technologies and materials development. Professor Kitchin, with over 41,500 citations on Google Scholar, is recognized for his expertise in catalysis and machine learning. He has also engaged in public outreach, including AMAs on platforms like Reddit and maintains an active presence on X (formerly Twitter) under the handle @johnkitchin.
While detailed member listings are not publicly detailed on the group's website, the team comprises graduate students, postdoctoral researchers, and collaborators who contribute to ongoing projects. Collaborations extend beyond CMU, including partnerships with institutions like Virginia Tech and MIT.
The Kitchin Group's research portfolio emphasizes computational and data-driven methodologies. Core areas include:
Recent publications highlight these focuses, including "Multiscale Perturbation Methods for Dynamic/Programmable Catalysis" (October 2025), "How Electrolyte pH Affects the Oxygen Reduction Reaction" (October 2025), and "Uncertainty Quantification in Graph Neural Networks With Shallow Ensembles" (September 2025). The group also explores programmable catalysis and mapping uncertainties using advanced programming techniques.
Professor Kitchin's work intersects machine learning, data science, and scientific programming, often applied to real-world challenges in science and engineering. This has led to over 176 publications and collaborations with industry and academia.
Agentic science represents a pivotal area of interest for the Kitchin Group, marking a shift toward AI-driven, autonomous scientific inquiry. In a September 2025 commentary published in Nature Machine Intelligence, co-authored by Hongliang Xin (Virginia Tech), John R. Kitchin, and Heather J. Kulik (MIT), the concept is defined as a paradigm where AI agents semi-autonomously conduct research by reasoning, planning experiments, and interacting with tools and equipment.
This builds on historical scientific eras—from empirical observations to computational modeling—and positions agentic AI as the next frontier. AI agents, powered by large language models, process multimodal data (text, images, structured information) to execute tasks independently. Examples include Coscientist, which interprets natural language and operates lab equipment, and A-Lab, an autonomous materials synthesis facility.
Kitchin argues that agentic science enables "self-driving labs," where AI acts as the decision-making "brain" amid robotics and cloud-based automation. This approach accelerates discovery in fields like environmental science and health, addressing limitations of traditional methods. Challenges include AI hallucinations, sensitivity to phrasing, lack of standardized evaluations, and resource demands, necessitating human oversight and ethical safeguards.
The paradigm fosters human-AI partnerships, improving reproducibility by analyzing vast literature and identifying research gaps. Kitchin emphasizes cross-institutional collaboration to avoid siloed development and ensure AI enhances scientific synergies. Recent advancements, such as AI's ability to search literature and plan experiments, underscore the timeliness of this work.
The Kitchin Group exemplifies innovative research at the intersection of chemical engineering and AI, with agentic science poised to revolutionize scientific methodologies. By advancing autonomous AI in labs, the group addresses pressing global challenges while emphasizing responsible integration. Future efforts will likely expand on self-driving labs, enhancing efficiency and discovery. For further details, visit the official website at https://kitchingroup.cheme.cmu.edu/.