Cognitive systems engineering (CSE) is aninterdisciplinary field that examines the intersection of people, work, and technology, with a particular focus onsafety-critical systems. The central tenet of CSE is to treat collections of people and technologies as a single unified entity—called ajoint cognitive system (JCS)—capable of performing cognitive work rather than as separate human and technological components. The field was formally established in the early 1980s byErik Hollnagel andDavid Woods.[1]
Unlikecognitive engineering, which primarily applies cognitive science to design technological systems that support user cognition, CSE takes a more holistic approach by analyzing how cognition is distributed across entire work systems. This perspective emphasizes understanding the functional relationships between humans and technology in complex operational environments such asair traffic control, medical systems,nuclear power plants, and other high-risk contexts.[2]
CSE draws on theoretical foundations from multiple disciplines includingcognitive psychology,cognitive anthropology,systems theory, andecological psychology. Key intellectual influences includeEdwin Hutchins'sdistributed cognition,James Gibson'secological theory of visual perception,Ulric Neisser'sperceptual cycle, andWilliam Clancey'ssituated cognition.[3] The field has also been shaped byJens Rasmussen's work onhuman error and abstraction hierarchy.
Methodologically, CSE employs techniques such as cognitive task analysis,[4]cognitive work analysis,[5] andwork domain analysis to understand how cognition is distributed across human and technological agents. These approaches focus on identifying system constraints and designing forresilience rather than merely preventing errors.
Cognitive systems engineering emerged in the wake of theThree Mile Island (TMI) accident.[6] At the time, existing theories about safety were unable to explain how the operators at TMI could be confused about what was actually happening inside of the plant.[7]
Following the accident,Jens Rasmussen did early research on cognitive aspects of nuclear power plant control rooms.[8] This work influenced a generation of researchers who would later come to be associated with cognitive systems engineering, includingMorten Lind,Erik Hollnagel, andDavid Woods.[6]
Following the publication of a textbook on cognitive systems engineering byKim Vicente in 1999 the techniques employed to establish acognitive work analysis (CWA) were used to aid the design of any kind of system were humans have to interact with technology. The tools outlined by Vicente were not tried and tested, and there are few if any published accounts of the five phases of analysis being implemented.[9]
The term "cognitive systems engineering" was introduced in a 1983 paper by Hollnagel and Woods.[1]
Although the termcognitive engineering had already been introduced byDon Norman, Hollnagel and Woods deliberately introduced new terminology. They were unhappy with the framing of the termcognitive engineering, which they felt focused too much on improving the interaction between humans and computers, through the application of cognitive science. Instead, Hollnagel and Woods wished to emphasize a shift in focus from human-computer interaction tojoint cognitive systems as the unit of analysis.[10]
Despite the intention by Hollnagel and Woods to distinguishcognitive engineering fromcognitive systems engineering, some researchers continue to use the two terms interchangeably.[11]
As mentioned in theOrigins section above, one of the key tenets of cognitive systems engineering is that the base unit of analysis is thejoint cognitive system. Instead of viewing cognitive tasks as being done only by individuals, CSE views cognitive work as being accomplished by a collection of people coordinating with each other and using technology to jointly perform cognitive work as a system.[1]
CSE researchers focus their studies on workin situ, as opposed to studying how work is done in controlled laboratory environments.[12] This research approach, known asmacrocognition, is similar to the one taken bynaturalistic decision-making. Examples of studies of work done in context include Julian Orr's ethnographic studies of copy machine technicians,[13]Lucy Suchman's ethnographic studies of how people use photocopiers,[14]Diane Vaughan's study of engineering work at NASA in the wake of theSpace Shuttle Challenger disaster,[15] and Scott Snook's study of military work in the wake of the1994 Black Hawk shootdown incident.[16]
A general thread that runs through cognitive systems engineering research is the question of how to design joint cognitive systems that can deal effectively with complexity, including common patterns in how such systems can fail to deal effectively with complexity.[17][12][18][19]
As mentioned in the Origins section above, CSE researchers were influenced by TMI. One specific application of coping with complexity is the work that human operators must do when they are supervising a process such as nuclear power plant, and they must then deal with a problem that arises. This work is sometimes known as anomaly response[12][20] or dynamic fault management.[21] This type of work often involves uncertainty, quickly changing conditions, and risk tradeoffs in deciding what remediation actions to take.
Because joint cognitive systems involve multiple agents that must work together to complete cognitive tasks, coordination is another topic of interest in CSE. One specific example is the notion ofcommon ground[22] and its implications for building software that can contribute effectively as agents in a joint cognitive system.[23]
CSE researchers study how people use technology to support cognitive work and coordinate this work across multiple people. Examples of such cognitive artifacts, which have been studied by researchers, include "the bed book" used in intensive care units,[24] "voice loops" used in space operations,[25] "speed bugs" used in aviation,[26] drawings and sketches in engineering work,[27] and the various tools used in marine navigation.[28]
Of particular interest to CSE researchers is how computer-based tools influence joint cognitive work,[29] in particular the impact of automation,[30] and computerized interfaces used by system operators.[31]
{{cite book}}:Check|isbn= value: checksum (help){{cite book}}: CS1 maint: location missing publisher (link){{cite book}}: CS1 maint: multiple names: authors list (link){{cite journal}}:Cite journal requires|journal= (help){{citation}}: CS1 maint: work parameter with ISBN (link){{citation}}: CS1 maint: work parameter with ISBN (link)