Carnegie Mellon University
Advanced Search   
  Look in
       Name    Email
       Former RI Members 
Sarosh Talukdar
Professor, RI/ECE
No longer a member of RI.
Personal Homepage
Research Interests

My research interests are in the areas of distributed problem solving, organizational theory, concurrent design (also called simultaneous engineering) and real-time control. I am working on the following projects:

Organizations and Super-Agents (A-Teams): An organization is a scheme by which agents, such as birds, fish and humans, combine to form super-agents, such as flocks of birds, schools of fish and corporations. The focus of this project is on organizations for robotic and computer-based agents. The objective is to develop organizations that are effective, so agents can cooperate in solving problems, and open, so new agents can be added whenever they are needed. Progress includes the creation of a set of organizational structures called Asynchronous Teams (A-Teams) in which large numbers of autonomous agents work in parallel, like ants in an ant colony. Application areas include the traveling salesman problem, diagnosis, global optimization, scheduling, and the design of reconfigurable robots.

Computer Aided Simultaneous Engineering (CASE). The objective of this project is to develop techniques for including downstream concerns, such as manufacturability and maintainability, in the early, conceptual stages of the design of an artifact. Key features of our approach include the development of constraint languages (schemes for articulating and enforcing design requirements), intelligent, autonomous critics (programs that can understand and criticize the decisions made by designers) and conflict resolvers (programs that can transform criticisms into suggestions for improvement). A number of different application areas are being used to drive the research, including automobile parts, electric circuits and buildings.

Contingency Theory: The quality of service provided by large systems, such as electric power networks, is threatened by contingencies, such as random failures and operating errors. The objective of this project is to develop strategies for reducing the ill effects of contingencies. The two possibilities are prevention and correction. Both are being considered and have been formulated as global optimization problems. Cooperative techniques for solving such problems are under investigation.