Advances such as commercial jetliners flown by a single pilot, astronauts on far-earth missions with minimal ground support, or semi-autonomous cars navigating urban roadways will all require ever more sophisticated automation. Yet while automation has enabled many of the advances in our lives, it remains brittle, hard to understand, and requires constant human oversight to maintain safety and performance. In many situations, the close coupling between humans and increasingly autonomous systems exacerbates human factors’ issues. Our work in human-autonomy teaming leverages advances in machine intelligence to create automation that behaves like a teammate – able to adapt to situations as they arise, explain what it is doing, and interact with humans in an intuitive way. Human-autonomy teams that are not designed with consideration of human needs and limitations can result in serious consequences that can compromise safety, including confusion, loss of situation awareness, cognitive overload, or misalignment of trust in automation. Our work addresses the fundamental research areas needed for human-autonomy teaming to fulfill its promise of operating as a teammate, including: adaptive systems, interaction paradigms, and the science of teams. Automation needs to adapt to the current situation without constant human direction. Human-autonomy collaboration requires richer methods of interaction, with a social element not found in traditional interaction with automation. Human-autonomy teams must support both task and team skills (e.g. coordination) to be successful. Additionally, our work has expanded to bring human factors methods to research projects that empower individuals and communities to make better decisions and encourage more inclusion in STEM.

The goal is to develop design principals and practical guidelines to foster the development of collaborative systems that support human activity in complex domains. Application areas include adaptive automation and adaptive interfaces, human-robotic interaction, interactive learning environments, and decision-support systems.