Ken Funk

Flight deck automation and task management

The purpose of the paper is to show that recent flight deck automation human factors research suggests that attention allocation or task management is a critical safety issue in advanced technology aircraft, to relate that finding to task management research, and to suggest a course for future research to address that issue.

The AgendaManager: A Knowledge-Based System to Facilitate the Management of Flight Deck Activities

Cockpit Task Management (CTM) is the process by which pilots selectively attend to tasks in such a way as to achieve their mission goal. Through our research we have found that CTM is a significant factor in flight safety, at least partly accounting for a substantial number of aircraft incidents and accidents. We developed an experimental knowledge-based system called the AgendaManager to facilitate Agenda Management (a superset of CTM) and demonstrated its superiority to a conventional crew monitoring and alerting system in a controlled evaluation study. The success of the AgendaManager is attributable not to its use of artificial intelligence technology. Rather, it is effective because it was developed using a sound human factors research and development approach. This approach and its application in AgendaManager development are the topics of this paper.

Agent-based pilot-vehicle interfaces: concept and prototype

The task support system (TSS), an integrative pilot-vehicle interface that consists of a collection of active software units called agents, is discussed. System agents represent aircraft and environmental systems and subsystems, providing databases and control interfaces to support the activities of task agents. Task agents represent tasks the pilot must accomplish in the mission. They assist the pilot by recognizing when the tasks should be started, configuring the cockpit for the tasks, alerting the pilot to potential errors, making recommendations on how to complete tasks, and automating tasks in certain cases. The TSS was evaluated in a simulation experiment as part of a prototype avionics system. Results of the experiment indicate that the TSS is effective in improving pilot performance and also was well liked by the pilots who tested it. >

Management of multiple tasks: cockpit task management errors

It is argued that a flightcrew has not only to perform tasks, but to manage them as well, and that proper task management can be crucial to the successful completion of the flight mission. From this perspective, a framework for cockpit task management (CTM) has been proposed. Based on this framework, a set of CTM errors is being derived from aircraft accident/incident reports and flight simulator experiments. The results of a preliminary analysis of 14 air accident reports are presented. It is found that not correctly initiating a task contributes to over one third of the errors in these accidents. An initial attempt to derive the cause of such errors indicates that the pilots knowledge of the aircraft or the procedures is usually limited in those tasks when abnormal conditions occurred. A preliminary set of specifications for a pilot-vehicle interface is proposed on the basis of the preliminary results.<<ETX>>

A Methodology to Identify Systemic Vulnerabilities to Human Error in the Operating Room

A methodology to identify systemic vulnerabilities to human error in surgical procedures was developed and applied to the initial stages of laparoscopic cholecystectomy. A database of generic human tasks and errors was developed and applied to four task descriptions developed from an IDEF0 model of the process. Over 30 vulnerabilities to human error were identified and prioritized.

Task Prioritization Factors: Two Part-Task Simulator Studies

Cockpit task management (CTM) is the initiation, monitoring, prioritization, execution, and termination of multiple, concurrent tasks by flight crews. In this research, we used 2 part-task simulator studies to elicit from pilots the factors actually used in task prioritization. Analysis resulted in the identification of 12 specific factors that the pilots reported as affecting task prioritization. From the results of these and related studies, we developed a hypothetical framework and model of task prioritization in which procedural consistency, importance to flight safety, and the salience of task-related stimuli are the primary factors that affect task prioritization.

A Functional Model of Flightdeck Agenda Management

Our research represents an effort to understand and facilitate the management of flightdeck activities by pilots. We developed a preliminary, normative theory of Cockpit Task Management (CTM) and from it defined an error taxonomy. Based on analyses using this error taxonomy we found CTM errors in 76 (23 per cent) of 324 aircraft accident reports and 231 (49 per cent) of 470 aircraft incident reports. Concluding that CTM is a significant factor in flight safety and recognizing the need to broaden as well as refine the concept, we developed a model of Agenda Management, which includes management not only of tasks, but goals, functions, actor assignments, and resource allocations as well. Major components of the functional model include maintaining situation awareness, managing goals (recognizing, inferring, and prioritizing), managing functions (activating, assessing status, and prioritizing), assigning actors (pilots and flightdeck automation) to functions, and allocating resources (such as displays and controls) to functions.

Agent-based aids to facilitate cockpit task management

In todays highly automated aircraft, the role of the pilot has changed from an airplane controller to a system manager. As a system manager in a cockpit, todays pilot is in charge of a supervisory activity we call cockpit task management (CTM). CTM activities include the initiation, assessment, prioritization, execution, and termination of tasks. This paper describes our past and ongoing efforts to understand CTM and to facilitate it through the use of agent-based, computational aids.

Use of Modeling to Identify Vulnerabilities to Human Error in Laparoscopy

This article describes an exercise to investigate the utility of modeling and human factors analysis in understanding surgical processes and their vulnerabilities to medical error. A formal method to identify error vulnerabilities was developed and applied to a test case of Veress needle insertion during closed laparoscopy. A team of 2 surgeons, a medical assistant, and 3 engineers used hierarchical task analysis and Integrated DEFinition language 0 (IDEF0) modeling to create rich models of the processes used in initial port creation. Using terminology from a standardized human performance database, detailed task descriptions were written for 4 tasks executed in the process of inserting the Veress needle. Key terms from the descriptions were used to extract from the database generic errors that could occur. Task descriptions with potential errors were translated back into surgical terminology. Referring to the process models and task descriptions, the team used a modified failure modes and effects analysis (FMEA) to consider each potential error for its probability of occurrence, its consequences if it should occur and be undetected, and its probability of detection. The resulting likely and consequential errors were prioritized for intervention. A literature-based validation study confirmed the significance of the top error vulnerabilities identified using the method. Ongoing work includes design and evaluation of procedures to correct the identified vulnerabilities and improvements to the modeling and vulnerability identification methods.

Human factors issues of flight deck automation

It is widely acknowledged that commercial transport aircraft automation has improved the economy of airliners, and that accident rates for advanced technology commercial aircraft are lower than those of comparable conventional technology aircraft. Nevertheless, criticism of the human factors of modern flight deck automation is common among pilots and other aviation professionals (e.g., Billings, 1997). Until recently, though, there was no comprehensive list of flight deck automation human factors issues, much less a comprehensive summary of evidence (data and other reasonably objective information) related to those issues. As a result, avionics designers have been somewhat in the dark about the usability of the equipment they have created to help improve the efficiency and safety of commercial air transportation. This paper summarizes a study we conducted to identify and compile evidence related to flight deck automation human factors issues. To be useful to those individuals directly involved in the development of future automation systems, the paper focuses primarily on those issues related to automation design.