Our society has greater need than ever before for people to collaborate effectively, whether they be teams of employees, patients and their doctors, large virtual communities, neighborhoods or family groups. Collaboration is involved in many societal problems including energy use, law enforcement, disaster relief, and scientific research. Researchers in human-computer interaction and computer-supported cooperative work have an impressive record of developing technology to support interpersonal communication and group interaction in a range of problem areas such as medical team coordination, disruption of family schedules, multi-tasking and interruptions in the workplace, energy use in households, collaborative intelligence analysis, virtual organization, and collaborative learning. Despite such efforts, the deep structure of collaboration is not well understood. The goal of this project is to acquire instrumentation that will enable the PI and her colleagues to leverage advances in other disciplines such as neuroscience and engineering to advance our knowledge of how details of brain and behavior are related to the underlying meaning, processes, and outcomes of collaboration, that is, its deep structure.
Intellectual Merit The requested research infrastructure will allow future projects to include a focus on the fine details of human behavior, including brain responses, facial expressions, body positions, gestures, and movements. The PI argues that in-depth analysis of behavior is necessary for understanding the mechanics of face-to-face interaction, for identifying those aspects of face-to-face interaction that are key to successful remote collaboration, for increasing the benefits of social media online, and for programming intelligent collaborative systems. Availability for interaction in the home environment, for example, might be indicated by a combination of posture and facial expression; detailed knowledge of these indicators could then help us design systems that deliver notifications to families at appropriate times. Similarly, remote experts collaborating around shared visualizations may benefit from camera systems that understand when partners have a shared understanding, as evidenced by their facial expressions and movements. Human-robot interaction might be refined by a fine-grained analysis of human-human interactions such as handoffs and greetings. And home health care and remote health delivery could be improved with better feedback about patient emotion and cognition during and after consultations. The infrastructure to be acquired with this funding will provide a platform for exploring ways to achieve these goals.
Broader Impact The ability to understand the deep structures underlying collaboration has the potential to radically transform science and engineering in general, by furthering our ability to analyze contributors to effective team performance across many domains, and by laying the foundation for new collaborative interfaces that support interaction through gesture, facial expression, eye gaze and brain activity. More specifically, there is a potential for direct broader impact in the domains of health and education. In the domain of health, the new techniques will enhance ongoing work at the PI's institution on chronic illness management and home-health care, supporting better collaboration between parents and children with chronic illnesses, and patients and therapists through high fidelity patient observation. In education, computer-supported and collaborative learning will advance by computer tutors understanding student attention and knowledge sharing in groups. The instrumentation will also have local impact in the CMU HCI Institute, allowing faculty and student projects to include a focus on understanding the dynamics of collaboration and on the development of new tools to enhance people's ability to collaborate in diverse settings.
