This project will develop a sensor-based real-time feedback control scheme for time-optimal and safe evacuation of people form large complexes with several exits, entrances, elevators, and stairs in an emergency scenario such as a major conflagration. The control scheme will interpret, use, and synthesize the data collected from sensors to formulate a feedback control command to be delivered through speakers, TV monitors or lighted matrix displays installed at strategic locations to channel an organized flow of entrapped people to safe locations in shortest time. The feedback control system will employ temperature sensors, smoke sensors, video cameras, and other people counting devices. It will assimilate the sensor data with the physical characteristics of the building considering the number of people trapped and the availability and capacities of various egress locations. This evacuation system can also be used for crises management of crowds in other chemically or biologically initiated emergencies, if appropriate sensor data are collected and effective control actuation delivery mechanisms are used. Software with animation features will be developed to visually demonstrate on computer screens the effectiveness of the proposed scheme under simulated fire emergency scenarios.
This project is supported under Sensors Initiative NSF 04-522.
