Brown archaeologist to co-lead AI project to advance understanding of ancient Andean empires Schmidt Sciences awarded $732,574 to a research team including Brown University archaeologist Parker VanValkenburgh and Vanderbilt University researchers Steven Wernke and Yuankai Huo for a project titled "Scaling Archaeological Vision: Imperial Landscapes of the Andes through AI-Assisted Survey." The team will build DeepAndes Workbench, an AI platform that lets archaeologists query an agent with images and verbal prompts to identify archaeological features and produce what is expected to be the first continental-scale maps of how the Wari, Inka and Spanish empires standardized, adapted and layered their built environments. The grant extends the team's earlier GeoPACHA geospatial platform and involves roughly 15 collaborators across the United States, Peru, Chile and Australia plus researchers at Oak Ridge National Laboratory, with Schmidt Sciences also providing in-kind supercomputing resources. PROVIDENCE, R.I. Brown University — With a new grant https://www.schmidtsciences.org/havi-2026-announcement/ from Schmidt Sciences, a research team that includes Brown University scholar Parker VanValkenburgh will use data science and AI techniques to advance archaeologists’ understanding of three ancient Andean empires. VanValkenburgh, an associate professor of anthropology and of archaeology and the ancient world at Brown, and Vanderbilt University researchers Steven Wernke and Yuankai Huo were awarded $732,574 for their project, “Scaling Archaeological Vision: Imperial Landscapes of the Andes through AI-Assisted Survey.” The work builds on more than a decade of previous research. The researchers will develop DeepAndes Workbench, an AI platform through which archaeologists and other scholars will be able to query an agent with images and verbal prompts to distinguish specific types of archaeological features. They say the tool will advance understanding of how three ancient Andean empires — Wari, Inka and Spanish — adapted to local conditions and successively influenced one another. “We’re thinking of DeepAndes Workbench like an image search engine for archaeological sites — a tool that will allow our colleagues to look for distinctive kinds of buildings and features — many of them built by the Wari, the Inka and the Spanish,” VanValkenburgh said. “Expert users will then provide feedback on the results they get back, which will in turn help us improve the models.” The result is expected to be the first set of continental-scale maps of how empires standardized, adapted and layered their built environments, he said. Researchers will be able to use the tool to detect specific features of, for example, Wari ceremonial buildings from the eighth century or Inka imperial buildings from the 15th century. “What you’ll get is not some all-seeing eye that knows everything there is to know about Andean archaeology, but a tool that researchers can use to extend their vision, which in turn models the kind of disaggregated, diverse and question-driven scholarly community that tends to exist in the humanities,” VanValkenburgh said. Previously, the team developed the Geospatial Platform for Andean Culture, History and Archaeology https://geopacha.org/ , or GeoPACHA, which uses satellite imagery and AI to map and study archaeological sites across the Andes. The new grant supports continued activities of GeoPACHA toward the development of DeepAndes Workbench. VanValkenburgh, Wernke and Huo are partnering with a team of approximately 15 collaborators across the United States, Peru, Chile and Australia, as well as researchers at the Oak Ridge National Laboratory in Tennessee, which is part of the U.S. Department of Energy and conducts research in areas including supercomputing. Schmidt Sciences is also providing in-kind supercomputing resources.