Background
Background |
|
| Synopsis: | This page defines key terminology used later in our description Cockpit Task Management (CTM). | ||
| Keywords: | system, behavior, goal function, task, cockpit task management | ||
| Author: | |||
| Ken Funk | <funkk@engr.orst.edu> | Department of Industrial and Manufacturing Engineering, Oregon State University, Corvallis, Oregon, USA | |
| Last Update: | 13 July 1999 | This is a Work in Progress and its contents are subject to continual revision. | |
Formally speaking, a theory is a collection of statements about some domain. These statements contain terms that are used to denote things and relationships that are considered to be important elements of the domain. For the theory to be sound and, of equal or greater importance, for it to be useful in analysis and design, these terms must be clearly defined. Terms essential to a theory of CTM are defined next.
A dynamic system is an entity that may be described in terms of input, output, and state. Input is matter, energy, or information flowing into the system. Output is the flow of matter, energy, or information out of the system. State is the set of system attributes at a given time. In addition, state is a compact representation of the history of the system that, with input given, makes possible the prediction of future outputs and states (Padulo & Arbib, 1974).
Two systems that are connected by inputs and outputs form a more complex system called a supersystem. If a system is formed from simpler systems through input output connections, the simpler systems are called subsystems. For example, an aircraft system can be partly defined as a collection of pilot, autopilot, airframe, and engine subsystems.
Note that this is "relative" terminology because whether something is called a system, a subsystem, or a supersystem depends on the analyst's perspective. For example, if the aircraft is considered a system, then the autopilot is a subsystem. On the other hand, if the analyst is primarily concerned with the autopilot, then the autopilot is a system, the aircraft is a supersystem, and the altitude hold circuitry in the autopilot is a subsystem. When using this terminology, the analyst must be careful to identify his or her purpose and frame of reference.
A system behavior is a discrete sequence or a continuous series of system input, state, and output values over a time interval. For example, given a system composed of airframe and engine subsystems, a behavior could be defined as time series of throttle setting (input), altitude (state), and sound pressure level (output) values. A system exhibits a behavior if observed values of input, state, and output values match those of the behavior.
An event is a set of system behaviors in which some state component changes in a significant way at the end of the time interval. For example, the event reach 10,000 ft consists of a set of aircraft behaviors, each ending with an altitude value of 10,000 ft.
A goal for a system is defined by a set of desired behaviors. If one of the behaviors is exhibited by the system, the goal is achieved, otherwise the goal is not achieved.
A goal has an initial event that defines the conditions under which the goal becomes relevant. For example, a typical flight path consists of a series of waypoints, which are geographical points along the route serving as intermediate destinations. So a goal to be at Waypoint 8 is relevant only after the initial event arrive at Waypoint 7 has occurred.
A subgoal of a goal is a set of behaviors consistent with those of the goal, but restricted in time and/or in scope. For example, a single goal to approach the destination airport and arrive at landing position (prior to final approach) could be decomposed into several subgoals: cleared to approach waypoint, at approach waypoint, approach flaps, approach power, and approach speed.
A goal and all of its subgoals form a hierarchy with the goal at the apex. The topmost goal for a flight mission will be referred to as the mission goal.
Goal priority reflects an ordering of a set of goals according to the relative importance or urgency assigned to them by the flight crew. More important or urgent goals have higher priorities. For example, a goal to remain clear of terrain and other aircraft established to maintain the safety of the aircraft and its passengers is clearly more important than a goal to avoid sudden maneuvers established for passenger comfort. The first goal should have a higher priority than the second.
This description of goal priority is a preliminary and highly simplified one. It does not accurately reflect the complex and dynamic nature of priority assignment which must take into account compliance with air traffic control directives, Federal Aviation Administration regulations, and company policies, to name just a few considerations. It reflects a minimal definition of goal priority which must be expanded as the theory is developed.
A task is a process that is completed to cause a system to achieve a goal. A task involves the behaviors of one or more secondary systems or subsystems necessary in order to produce inputs to the primary system to achieve the goal. For example, for the goal to arrive at Waypoint 7, there must be a fly to Waypoint 7 task. The pilot, the primary flight controls, the cockpit displays, the hydraulic system, and the engines are just a few of the secondary systems required to complete the fly to Waypoint 7 task to achieve the goal for the primary system (the aircraft) to arrive at Waypoint 7. These secondary systems are called resources. Stated another way, tasks require resources to achieve goals.
A task has state. Initially, a task is latent. When the initial event of its goal is imminent, the task becomes pending. When the initial event occurs, the task becomes active. A task becomes active in progress when resources are allocated to it to achieve the goal (i.e., while it is being executed). If the task has been in this state but resources are deallocated from it and execution ceases, the task returns to the active state. A task may be terminated if its goal is achieved, if the goal is unachievable, or if the goal becomes irrelevant. In the case of an unsuccessful termination, the task may be considered to be aborted. Further state decomposition is possible and perhaps desirable, but the set of states just described is satisfactory for the preliminary theory to be presented later in this document.
The goal to approach the airport and arrive at landing position was decomposed into cleared to approach waypoint and at approach waypoint subgoals. Similarly, an approach task could be decomposed into get approach clearance and fly to approach waypoint subtasks.
An agenda is a hierarchy of tasks to be completed during a mission. Each task is defined to achieve a specific goal and should become active when the goal's initial event occurs.
When an initial event occurs, the corresponding task becomes active. Two tasks that are simultaneously active are called concurrent tasks.
Resource-Limited PerformanceExecuting a task involves the coordinated behaviors of one or more systems or subsystems called resources. Certain resources are required to complete each task, and if the resources are not available, the task cannot be completed satisfactorily and the goal cannot be achieved.
A variety of resources are required for cockpit tasks. Equipment resources include autopilots, radios, displays, and controls. Human resources include the captain, first officer, and flight engineer. Because resources are systems, they can be decomposed into simpler subsystems. Human resources can be decomposed into personal sensory, motor, and cognitive resources. Cognitive resources can be further decomposed into the verbal and spatial resources identified and studied by Wickens and his colleagues at the University of Illinois (Wickens, 1984; Wickens & Liu, 1988).
Because two concurrent tasks may require the same resources, this poses a potential problem. Behaviors of necessary resources that are compatible with achieving one goal may be incompatible with achieving another goal, and the performance of one or more of the tasks may suffer. That is, task performance is limited by resource availability. With resources like displays or hands and feet, this is obvious. But it is also true for cognitive resources (Navon & Gopher, 1979; Wickens, 1984). A situation in which task resource requirements exceed resource availability is called a task conflict.
For example, in the situation described above, if air traffic control clearance to an approach waypoint is obtained the set and maintain approach power task would become active. Assume that this task requires a multifunction display resource on which an engine display format must be shown. Suppose that now a primary electrical system failure event occurs and a subtask to diagnose/correct the electrical system becomes active. Assume that this subtask requires an electrical system display format on the same display resource. If the two formats cannot be displayed simultaneously, a resource shortage resulting in a task conflict exists.
Even if two displays are available to complete both of these tasks simultaneously, there still might be a task conflict due to cognitive resource limitations. Assuming for the purpose of this illustration that no other crew member is available to assist the pilot in completing these two tasks, he or she may lack sufficient cognitive resources to attend to both of them simultaneously. This might result in errors in completing one or both of the tasks.
Listed below, most recent first, are changes made to this page since its creation.