Co-PI: Vijay Kumar, U. Pennsylvania Co-PI: Sanjiv Singh, CMU
This collaborative ITR project addresses the development of proactive networks of sensors and robots that perceive their environment and respond to it, anticipating information needs by the network and by users of the network, repositioning and organizing themselves to best acquire and deliver the information. Such networked systems combine the most advanced concepts in perception, communication, and control to create computational systems capable of interacting in meaningful ways with the physical environment, extending individual capabilities of each component and user to encompass a much wider area and range of data. The PIs envision a physical analog to the Internet-- networks of computers that can actively sense, physically interact with, and reason about the world. While the Internet allows transparent access to information already online, this research will extend the paradigm by allowing users to "google" for physical information, setting into motion robots and sensors that team together to acquire information and act on it.
Intellectual Merit
The proposed research program will be make significant contributions to networked multi-agent systems: control, self-organization, adaptation, and perception. The work will focus on: (1) Control for communication and sensing: the control of robotic agents to maintain communication links or establish new ones, while obtaining the required sensory information and tracking sources; (2) Communication for sensing and perception: the fusion of information from heterogeneous sensors over the network, providing the required information for each agent to plan and control its mobility and providing remotely located human rescue workers with information through immersive displays; and (3) Communication networks for sensing and control: the grouping, scheduling and routing of nodes to adapt to changing, adverse conditions while maintaining guarantees for control of mobility and for sensor fusion and integration.
Broader Impact
The research will enhance national and homeland security by providing first responder with information about areas that are unsafe and hard to reach for humans in three ways. First, it will speed up the response time by helping in assessment, by augmenting human perception for command and control. Second, it will help in suppression and containment. Third, it will play a role in recovery, as in identification of victims and location of responding personnel. The collaboration with practitioners at the Allegheny Fire Training Academy will ensure that the research is grounded in the real world and has impact in the emergency response community.
