ESD logo
Site Map | Contact | Search

 

ESD Research Domains

 

Energy and Sustainability

 
  Extended Enterprises
 
  Health Care Delivery
 
  Critical Infrastructures
 

ESD Research Approaches

 

Humans and Technology

 
  Uncertainty and Dynamics
 
  Design and Implementation
 
  Networks and Flows
 
  Policy and Standards
 

 

 

Real-Time Predictive Human Supervisory Control Models of Team Collaboration

Complex systems are typically managed by difficult-to-supervise teams of human controllers. Feedback about interactions between team members, as well as with the system, may not be observable, and such critical collaboration factors as team knowledge and shared cognition are difficult to assess in real time.

nasa

 

NASA’s control room of the International Space Station exemplifies how human beings are increasingly required to work with multiple layers of technology.

Image courtesy of NASA

The goal of this project is to build models of team behaviors able not only to recognize the current state of a team supervising automation in real time, but also to predict future states of this team. Specifically, the team models are based upon the observation of behavioral patterns at both the individual and collective levels. A main contribution of this project will be to determine the robustness of the prediction of future team behaviors based on observing social patterns of collaboration. This project is therefore at the intersection between artificial intelligence and social sciences. G iven the prevalence of team interaction with many complex systems such as air traffic control, disaster first response, and military command and control, this research is relevant to numerous high-risk
critical systems.


Boussemart, Y., & M.L. Cummings, “Behavioral Recognition and Prediction of an Operator Supervising Multiple Heterogeneous Unmanned Vehicles,” H umans Operating U nmanned Systems `08, S eptember 3–4, 2008, B rest, F rance.

 
         
MIT SoE MIT Sloan School of Management MIT School of Science SHASS SA+P