Abstract:
|
In this presentation we overview a methodology for determining automated heating, ventilation and air conditioning (HVAC) and building personnel actions in response to chem/bio threats in the workplace. Models were developed and applied for counteracting threats based on distributed sensor data. The implementation of automatic system mitigations (e.g. HVAC airflow changes) was determined by modeling and analyzing key factors which include the impact to health and productivity of the hazard, the action decision, false positive and false negative rates, and trust in an automated system. Our models were evaluated on simulated scenarios using the National Institute of Standards and Technology multi-zone indoor airflow analysis computer program, CONTAM. Personnel actions (e.g. full evacuation) were implemented when airborne contaminant concentration level thresholds were exceeded. These thresholds were determined based on modeled notional effects of chemical and biological hazards for acute and chronic exposure limits. Our ultimate objective is to incorporate data streams into an automated decision model and create optimal policy based upon minimizing metrics such as cost and regret.
|