Federal Grant Award Summary

The National Science Foundation's Division of Computing and Communication Foundations awarded the University of Maryland, College Park a $666,666 Project Grant (CFDA 47.070: Computer and Information Science and Engineering) effective September 1, 2026, through August 31, 2029. This collaborative research initiative develops advanced Gaussian Process (GP) theory and algorithms capable of handling heterogeneous sensor networks, autonomous agents, and data sources—addressing a critical gap in current GP methodology that assumes homogeneous systems. The project delivers three primary research thrusts: (1) algorithms enabling autonomous systems with disparate movement capabilities, costs, and constraints to collaborate on efficient data collection; (2) sensing methods accommodating sensors with varying operational characteristics and measurement fidelity; and (3) techniques for reasoning about phenomena exhibiting inherent spatial or temporal heterogeneity. These deliverables support improved active learning frameworks for autonomous multi-sensor sampling missions.

The award advances practical mapping and decision-support capabilities by enabling efficient, cost-effective data acquisition through optimized sensor placement and timing decisions. By extending GP-based active learning to real-world heterogeneous systems—rather than idealized homogeneous environments—the research facilitates development of autonomous systems capable of generating accurate environmental maps with fewer overall samples, reducing the time and expense associated with field data collection. The work is conducted at the University of Maryland's College Park campus and contributes to the NSF's broader mission of supporting investigator-initiated computing research that accelerates discovery and innovation through advanced computational methodologies.