MOSAIC Lab
Manufacturing Operations: Simulation, Automation, Intelligence, and Connectivity
Industrial Engineering Program, Department of Mechanical Engineering
Molinaroli College of Engineering and Computing, University of South Carolina
Putting the pieces of the digital factory together.
The MOSAIC Lab studies how digital technologies integrate into working cyber-physical systems, and builds the infrastructure that lets manufacturers use modern Industry 4.0 tools and puts those tools in the hands of the operators and engineers who run the line. The lab's work covers the main aspects of industrial digitalization: the sensing and connectivity that gets data off a shop floor, the models and simulation techniques that keep a virtual operation synchronized with the physical one, the software that executes those models on live data, and the assistants and decision-support tools that let operators and managers use the results. The lab's methods favor open-source, easy-to-access technologies, designed to be within reach of small and medium manufacturers and research labs.
The lab is directed by Dr. Michail Katsigiannis and was established in 2026. It is also developing a physical testbed, an educational assembly line for Smart Manufacturing and Lean Manufacturing education and research.
Two funded PhD positions open for Spring or Fall 2027Research
The lab's research spans four areas, from the shop floor to the people who run it.
Modeling and Simulation for real-time operations
Traditional Modeling and Simulation techniques (Discrete-Event Simulation, Agent-Based Modeling, and their combination) were built for offline studies of a system before it exists or apart from its operation. The lab works on transforming these techniques to support real-time operations, and on integrating physics-based models with higher-level operational models into a single operational model of the factory.
Shop-floor sensing and connectivity
The lab designs the embedded systems, connectivity, and edge computing that get data off a shop floor and into a Digital Twin: product tracking, industrial robot and PLC integration, computer vision, and retrofits that bring legacy equipment online. The emphasis is on approaches a manufacturer without a dedicated IT staff could deploy.
Distributed Digital Twin architectures and engines
The lab develops distributed Digital Twin architectures and the engines that execute them on live data, and studies what it takes to deploy them at scale: interoperability with existing industrial systems, fidelity to the physical line, and security. The goal is Digital Twin infrastructure that runs without commercial simulation software and that manufacturers and research labs can adopt.
Intelligent applications for cyber-physical systems
The lab builds the applications that let people use a Digital Twin: locally hosted Large Language Model (LLM) assistants that answer questions about the factory from live Digital Twin data, classifiers that translate natural-language requests into safe cyber-physical actions, adaptive dashboards, and adaptive manufacturing decision-support systems that change what they show and recommend as the state of the operation changes.
Facility
Smart Manufacturing and Lean Systems teaching laboratory
The MOSAIC Lab is building an educational assembly line for Smart Manufacturing and Lean Manufacturing education and research in USC's Industrial Engineering program, proposed in 2026 and now in its first phase of development.
The design follows the Tiger Motors Lean Education Center at Auburn University, where Dr. Katsigiannis led the digitalization research under a Department of Defense program from 2023 to 2026. Tiger Motors is a 15-workstation mixed-model assembly line that builds automobiles from LEGO bricks at a 70-second takt time and runs in both Mass Production and Lean Manufacturing configurations. The USC line will reproduce that design and instrument it from the start: product tracking at every workstation, PLC and Andon automation, a live agent-based Digital Twin, and, in later phases, collaborative robots, machine vision, and extended-reality training.
The facility is built as a testbed that teaches students how to work in a modern industrial environment, demonstrates Industry 4.0 technologies to local small and medium manufacturers, and supports digitalization research. It will be the teaching lab for the IE program's production, simulation, and laboratory courses, a training site where South Carolina manufacturers can run a Lean line themselves, and the MOSAIC Lab's research testbed, producing real production data from a system that can be experimented on without disrupting a plant.
Development is phased. The assembly line and Lean tooling come first, followed by sensing, networking, and the first Digital Twin; automation, robotics, vision, and material handling follow as funding and space allow.
People
Dr. Michail Katsigiannis, Director
Assistant Professor, Industrial Engineering Program, Department of Mechanical Engineering, University of South Carolina
Ph.D. and M.S., Industrial and Systems Engineering, Auburn University. Integrated Master's (Dipl.-Eng.), Electrical and Computer Engineering, Technical University of Crete.
Michail Katsigiannis works on Digital Twins of manufacturing operations: how to model them with Agent-Based and hybrid simulation techniques, how to connect them to shop-floor equipment, and how to run them without commercial simulation software. His doctoral work at Auburn University's Interdisciplinary Center for Advanced Manufacturing Systems developed the Multi-Agent-Based Digital Twin architecture and its execution engine, and applied them to turn the Tiger Motors assembly line into a demonstration facility for Digital Manufacturing under a U.S. Army-funded program. Earlier work covered sensing systems for small and medium manufacturers, IoT-based monitoring of additive manufacturing, and multi-agent control of building energy systems. He was a Chapman Foundation Simulation Fellow and a Walt and Virginia Woltosz Fellow at Auburn.
mkatsigiannis@sc.edu · Room C113A, 300 Main St., Columbia, SC 29208 · Google Scholar · CV (PDF)
Students
Two funded PhD positions are open for Spring or Fall 2027. See Join the lab.
Join the lab
PhD students
Two funded PhD positions are open to start in Spring or Fall 2027. Each position carries a graduate assistantship (stipend and tuition support) for the first two years, with continued support through research funding expected for the duration of the program, contingent on satisfactory progress. Students apply to the Mechanical Engineering Ph.D. program and pursue Industrial Engineering research in the lab.
Students in the lab work across simulation, software, and hardware. A typical project involves building a model, connecting it to equipment or to a data stream, and getting it to run in real time, so applicants should be interested in building systems that combine software, hardware, and models, and in at least one of: Discrete-Event Simulation or Agent-Based Modeling, Digital Twins, industrial IoT and edge computing, machine learning and Large Language Model applications, or robotics and automation. Backgrounds in Industrial, Mechanical, Electrical, or Computer Engineering, or Computer Science, all fit. Programming experience (Python, Java, or similar), software development, DevOps, or networking helps, as does experience with AnyLogic, Simio, or PLCs; none of these is required.
The assembly-line facility will be built during the first cohort's time in the lab. Students who join early will help design and instrument it.
To express interest, email mkatsigiannis@sc.edu with a CV, unofficial transcripts, a statement of no more than one page on which research area interests you and why, and links to code, a simulation model, a thesis, or a project you built. For Spring 2027, send materials to the lab as early as possible (USC deadlines: October 1, 2026 international, October 15, 2026 domestic). For full consideration for Fall 2027, send materials to the lab by January 15, 2027. Contact the lab before applying formally; formal applications go through the Mechanical Engineering Ph.D. program (USC deadlines: May 1, 2027 international, May 15, 2027 domestic).
USC undergraduates and master's students
Students in the IE and ME programs who want research experience are welcome year-round. Work ranges from building simulation models to wiring sensors and writing dashboard code. Capstone teams can take on projects tied to the assembly-line build.
Selected publications
- Mykoniatis, K., & Katsigiannis, M. (2025). Digitalizing the automotive assembly supply chain using multi-agent based digital twins. In G. Rabadi & B. Soykan (Eds.), Optimizing supply chains through digital twins (pp. 197-220). Springer.
- Katsigiannis, M., & Mykoniatis, K. (2024). Real-time tracking of production in assembly operations using agent-based modeling and digital twin techniques. Proceedings of the 2024 Winter Simulation Conference, 3011-3022.
- Katsigiannis, M., Pantelidakis, M., & Mykoniatis, K. (2024). Assessing the transition from mass production to lean manufacturing using a hybrid simulation model of a LEGO automotive assembly line. International Journal of Lean Six Sigma, 15(2), 220-246.
- Katsigiannis, M., Evans, M., Osho, J., Pantelidakis, M., Bitencourt, J., & Mykoniatis, K. (2024). Empowering decentralized production: A distributed manufacturing system for additive manufacturing processes. Manufacturing Letters, 41 (NAMRC 52).
- Katsigiannis, M., & Mykoniatis, K. (2024). Enhancing industrial IoT with edge computing and computer vision: An analog gauge visual digitization approach. Manufacturing Letters, 41 (NAMRC 52).