Georgia Tech Researchers Build Dual AI Platforms to Transform How We Understand Complex Systems

Georgia Tech's Tech AI hub has developed two AI-driven platforms — research software engineer Jeremy Garcia's 3D Visualization Platform and Austin Weeks' Armada logistics platform — that let users explore complex systems from global supply chains down to individual warehouses and manufacturing plants. Armada maps warehouses, distribution centers and trucking routes across international, national, state and city scales to simulate scenarios and identify bottlenecks, while Garcia's platform turns static data into an explorable 3D space for systems such as chemical manufacturing lines and energy grids. Tech AI head of engineering Dave Brownell said the tools "help visualize and identify bottlenecks in large workflows" and remove most of the burden of building such platforms from scratch; both are already used in student learning, exploratory research and experimental modeling.

Georgia Tech researchers are advancing how people visualize, manage, and interact with some of the world’s most complex systems, from the national energy grid to large-scale logistics networks. Two new projects developed within Tech AI, the artificial intelligence AI hub at Georgia Tech, highlight this drive. Tech AI’s Research Software Engineers Jeremy Garcia and Austin Weeks, who work under Dave Brownell, Tech AI’s head of engineering, are creating AI-driven platforms designed to turn complexity into clarity while keeping ease of use at the forefront. Their goal is simple: Displaying a dataset should require minimal effort, whether the data represents a local model or global infrastructure. Both projects, Garcia’s 3D Visualization Platform and Weeks’ Armada , a logistics platform operating at global, national, state, and city levels, reflects Tech AI’s commitment to practical, research-backed solutions that are ready to be deployed across industry and academia. Both researchers picture a future where their platforms work together in a seamless way. A user could begin with Armada’s nationwide or international perspective, tracking supply chain flows and warehouse networks, and then zoom into a specific warehouse or plant to explore its internal operations through Garcia’s 3D Visualization Platform. Seeing Systems from Every Angle Garcia’s platform transforms static data into a fully explorable 3D space. With the Visualization Platform, users can interact with complex systems, from chemical manufacturing lines to energy grids, and uncover insights that would otherwise remain hidden in spreadsheets or charts. “You can zoom in on a single manufacturing plant or warehouse and watch its internal operations come alive,” Garcia said. “Once you can visualize it, you can innovate on it.” This micro-level view helps researchers, engineers, and students understand intricate processes that are often too complex to grasp otherwise and bridges academic study with real-world application. A Bird’s-Eye Perspective for Logistics While Garcia’s platform focuses on internal operations, Weeks’ Armada provides a macro view of logistics systems at almost any scale. Armada can visualize scenarios internationally, nationally, statewide, and down to individual cities. It maps warehouses, distribution centers, and trucking routes to create a dynamic and interactive model of large-scale supply chains. Armada enables organizations to simulate scenarios, identify bottlenecks, and optimize operations from a high-level perspective with minimal setup time. “You can see the entire network at once, and then drill into whatever matters most,” Weeks said. “It is powerful because organizations rarely get that kind of clarity.” Macro Meets Micro This integration would provide a new level of insight for industries such as retail, logistics, manufacturing, and robotics. It would allow organizations to understand and optimize complex systems from the broadest scale down to the smallest operational details. These platforms also reflect Georgia Tech’s identity as an AI-forward institution, where research is paired with practical and deployable solutions. Both tools are already supporting student learning, exploratory research, and experimental modeling. “These tools can be used in both industry and academia,” Brownell said. “They help visualize and identify bottlenecks in large workflows, and they showcase optimizations achieved through targeted research. With them, complex systems become much easier to understand, and the value produced by the underlying processes is more apparent. Platforms like these are often difficult or time-consuming to build from scratch, but ours remove most of that burden and make powerful visualization capabilities accessible to everyone.”