Operational Decision Making in Aluminium Smelters

Many computer systems incorporating artificial intelligence have been introduced for use in industry to assist in making decisions and controlling processes. However, decision making in a complex industrial plant, such as an aluminium smelter, involves psychologically related factors such as intuitive reasonings, operator response characteristics, perception of risk, and implication of rewards. While a significant body of work does exist on decision science, research concerning human interaction with process control systems is still at the development stage. The work reported here aims to meet the needs of the process industry by incorporating human factors and decision making strategies into computer programs such as a supervisory control system for aluminium smelters. A case study on the control of the level of the liquid electrolyte was carried out to firstly facilitate an understanding of the variables, including human factors, on process control. It was found that the availability of crushed solidified electrolyte material had a significant impact on the level of the liquid electrolyte, while the implementation of a supervisory control system had a certain impact, management and leadership styles also had a significant influence.