The main goal of this project is to increase the efficiency of and ease file pressure on air-traffic controllers by creating software intelligent agents to provide real-time assistance wifil their work. To build an intelligent agent we have first conducted a task analysis on the work of air traffic control in order to establish a model of the task. Two separate hours of air traffic control operations were video taped and analyzed in detail. One was a relatively normal hour whereas the other was during a heavy rainstorm that significantly affected the flow of air traffic. Interestingly, the main characteristic of the rainstorm hour was its nonroutineness, while the normal hour showed a consistent and regular interplay between the goals of efficiency and safety. Based on file outcome of file task analysis we can begin to assess and diagnose situations where air traffic controllers are more likely to experience cognitive overload. This model will be the reference for building an intelligent agent to airport ATC activities.