Download as pdf
Download as pdf
You are on page 1of 7
RR-JORBS NUWC-NPT Reprint Report 10,236 45 July 1983, Human Factors Engineering Methodologies in Submarine Combat System Concept-of-Operations Experiments M. F. Kanter F. J. O'Brien Combat Control Systems Department HOAL LIBRARY £0 WARFARE CENTER Naval Undersea Warfare Center Division Newport, Rhode Isiand Approved for public release; distribution és uniimited. Reprint of a paper presented at the Human Factors Society 33rd Annual Meeting, Octaber 1969, Denver, CO. PROCEEDINGS of the HUMAN FACTORS SOCIETY 33rd ANNUAL MEETING —1989 Human Factors Enginoering Methodologies in Submarine Combat Systems ‘Concept of Operations Experiments Mark B, Kanter and Frank J. O'Brien Naval Underwater Systems Center Newport, Rhode Island 02841-5047 ABSTRACT Concept of Operations Experiments (COOPEXs) are conducted at the Naval Underwater Systems Center to evaluate submarine combat system operability through structured walkthroughs of submarine missions in a full-scale replica of the combat system environment. Data were collected from one COOPEX for the purpose of piloting human factors engineering methodologies. Partial results based on a different COOPEX scenario are reported and compared. The data reduction and analysis procedures of digraph analysis, Q- analysis, multidimensional sealing and erew density were applied to assess combat sysiem information flow and configuration effectiveness, Results revealed the potential for significantly enhancing submarine combat system performance when applied to larger, more complex data sets, Plans for subsequent research are discussed. INTRODUCTION This paper discusses the development of data reduction and analysis methodologies for the human factors assessment of submarine combat, system operability. The combat system is expected to provide the Commanding Officer (CO) with the requisite data for decision making. Concept of Operations Experiments (COOPEXs) are eonducted at the Naval Underwater Systems Center in an effort to evaluate performance for postulated combat system configurations. Subject matter experts (ex-fleet personnel) exercise the configuration by participating in structured walkthroughs of submarine missions in a ful scale replica of the submarine combat control center, Results from the COOPEXs have led to hardware and software configuration modifications. However, the processes through which those results’ were obtained--the methodologies--have been limited to qualitative analysis (Wallin, 1988). These qualitative analyses have been conducted because quantitative analysis precedents are almost nonexistent or difficult to implement. Submarine command and control systems consist of combat system (CS) elements needed by the Commanding Officer and other attack center personnel to exercise command of and to control the CS assets required to conduct submarine warfare (Submarine Command and Control Systems, 1983). These subsystem elements and the personnel that utilize them come together in the submarine combat control and attack center (figure 1) visits the following locations/abjects "L combat control and attack center during a COOPEX scenario: + Tactical situation plotter-TSP Geographic plotter~GEOPLOT. + Interior communieations.-PHONE 1187 Figure 1. SSN-21 Combat Control and ‘Attack Center Sonar room--SONAR ‘avigation plotter--NAVPLOT Combat system display consoles (Operator Workstations)-CSDCs Junior-officer-of-the-deck/auxiliary command displays~ JOOD/ACD Officer-of-the-deck-OOD During a structured walkthrough individuals follow a script. ‘The Seript informs them of their tasks but not the procedures or the operational activities to cary out their tasks COOPEXs are not controlled experiments because the researchers do not have a priori knowledge of the subject matter experts’ operational activities. The opportunity exists to understand the operational activities through exploratory data Qualitative analysis of the fiscal year 1987 (FY87) COOPEX (Wallin, 1988) indicated the need to relocate various equipment to improve accessibility by key personnel and the requirement for a group display. ‘The results obtained were generated from human observation PROCEEDINGS of the HUMAN FACTORS SOCIETY 33rd ANNUAL MEETING —1989 indication of the subject's behavior during the data collection session. However, this information can become incomprehensible in a static representation as shown. ‘The data from the FY87 and FY88 COOPEXs were analyzed to determine the reliability of both the analysis procedures and results, The CO's movement behavior was anslyzed with respect to the following attack center locations: Figure 3. CO Trajectory for FY88 COOPEX ‘TSP. GEOPLOT PHO! SONAR NAVPLOT cscs JOODIACD oop Data collected from both COOPEXs were also analyzed with respect to the following mission phases: transiting, contact prosecution, battle stations, and firing point procedures, Analysis Procedures Representative _descriptive/inferential statistical and graphical exploratory data analyses were used for illustrating the potential benefits of the following quantitative mothodologies: + Process flow analysis~a systematic observation of task performance, Digraph analysis--graph_ theoretic techniques used to identify strong points and activity cycles from adjacency matrices of sequence data, Structural modeling--A method to understand the structural dynamics of man-system interactions. a) Q-analysis-a mathematical’ 1189 technique that structural in decision- graphical examines relationships making, Population Density Index (PDj)--a quantitative measure that accounts for crew density as a function of spatial orientation of the number of personnel and is scaled for the geometric area of interest. Multidimensional sealing-a multi- variate technique for determining an object space proportional to inter-object distances for the purpose of deter mining a configuration link analysis. » Flow Analysis Process flow analysis was used to describe the CO's behavior at selected time intervals of the scenario. The amount of distance traveled, activity type and activities per minute is recorded, Structural M . It is necessary in any configuration modeling to first understand the structural dynamics of the man-system interactions. Two approaches are used on sequence data. Elementary graph theory (digraphs) are useful to detect interesting patterns such as eyclic activity patterns and strongpoints (e.g, Harary et al., 1965) A second approach to structural modeling is provided through a little known modeling methodology. This methodology referred to a8 Polyhedral Dynamics Analysis (Atkin and Casti, 1982) is a complex theory applied to binary data Gefined at various hierarchical levels by “cover sets”. Polyhedral Dynamics Analysis has produced structural depictions of many diverse systems (from poems to football games and military command and control systems) not. otherwise amenable to quantification. The flexibility of PDA lies in the analysis of mathematical relations as opposed to functional analysis. Our use of PDA has been limited solely to the algorithm roferred to as Q-analysis (Johnson, 1988) which is essentially a single-level (nearest neighbor) hierarchical clustering technique assuming a similarity coefficient calculated by counting common descriptors applied to a one: level hierarchy, ‘Thus, distinct but related structural methodologies (graph theory and PDA) provide a complementary interpretation of system structure. Based upon the data presented, these structural analyses are limited. For example, PROCEEDINGS of ne HUMAN FACTORS SOCIETY 33rd ANNUAL MEETING—1989 inter object/location distance is ignored. Crowding is an_obvious problem in the small workspaces of a submarine. As an individual works more closely with others, the need for space is heightened, thereby causing increased. stress-related behavioral and physiological effects (McCain et al., 1976; Stokols et at., 1973). This demand for space has also been found to be evident during naval vessel operations (Dean et al., 1975). A mathematical model was constructed ta predict the amount of density expected in the 688 class submarine and the SSN- 21 (Seawolf) submarine. Population density is usually defined as an objective measure of the number of people per unit area of space (Baum and Epstein, 1975) Clearly this definition is limited beeause spatial orientation is disregarded. A measure of population density is propased that accounts for crew density as a function of spatial orientation of the number of personnel and is scaled for ihe geometric area of interest. ‘The Population Density Index (PDD is given as: pol = WaA/G @ where d is the average pairwise distance among nn people at locations (xj, yjii= 1, 2... n) within a defined geometric area of A square units, ie = Dano /(2) The expression for PDI assumes. that distance is proportional fo the square rot of area, A, (Rand Corporation, in Walker, et al., 1975). The value YAMA corresponds to the vertical or horizontal distance between any consecutive pair ‘f'n equally spaced points in the area, A Gtansen et al, 1953). Thus, the PDI statistic measures the verage distance between any two erew members as they are actually distributed sealed by the distance between adjacent crew members assuming that they position themselves Uniformly throughout the workspace. ‘The population density index was used as a measure Srerow density in the combat center workspace environment ‘Multidimensional Scaling, Multidimensional scaling procedures were used to determine a configuration based upon the CO's behavior (i.c., his frequency. of communications and interactions). MDS. provides a computationally efficient method for evaluating link data associated with a link analysis and has been applied to facilities layout 1190, (Biegel ot al., 1982; Tullis ot al., 1986). ‘The criteria for assessing relationships among CO system utilities were (1) frequency with whieh the CO shifted from one man/machine cbject to ‘another and. the degree of crowding (an alternative for crew stress) that existed while the CO transitioned between system objects and locations, @ Tullis, et. al. (1986) provides a useful technique for combining functional eriteria. ‘Two separate matrices were generated: F for CO transition frequency and P for population density. Individual similarities matrices were summed for each evaluative dimension--frequeney of interaotions and crew density SUMMARY OF RESULTS 1. Process analysis suggests that as each phase of the mission, and the mission as a whole, gains definition the CO experiences travel pattern changes. Possibly, as the mission reaches culmination, the amount of loading experienced by the CO may be increasing due to the type and amount of information processing needed to make accurate and timely decisions 2 Structural modeling via digraph analysis and Q-analysis provided insights into understanding the structure of information flow: Similar structures were discovered in each COOPEX implying a core structural component to CO behavior in different submarine scenarios, 3. The population density index (PDI) was introduced as a useful relative measure of personnel congestion in the combat center. During the different mission phases it was determined that the density of people increases as the mission evolves. Density changes due to increases in the number of personnel as well as their clustering in specific areas of the combat center are largely determined by the CO's activities. It was noted that, during battle stations the CO and the junior officer-of-the-deck had difficulty hearing and talking with other crew members because of the density of people at, the GEOPLOT. 4, The multidimensional scaling on combined criteria PDI and CO frequency of movement suggests that the combat center is organized into several groups of displays: the NAVPLOT/OOD; SONAR; and GEOPLOT, CO's displays, and the CSDC displays. The direct implication of this is that communication between these groups may be weak and that the CO, or other personnel, must move throughout PROCEEDINGS of the HUMAN FACTORS SOCIETY 33rd ANNUAL MEETING —1989 the combat center to access information. Reexamination of display organization to include fewer and larger systems of viewing envelopes (group displays) may be particularly relevant considering newer trends in automation and the reduction of operator-level manning. OF course, results are only suggestive. Meaningful inference will come only from a more deliberate examination of additional and more representative data sets. An important quality of these methods is the potential to examine system parameters that transcend fundamental spatial arrangement of system elements. Cluster analysis and muftidimensional scaling provide the means to examine dimensions of panel and display layout, as well as more abstract qualities of {ask analysis and information representation. ‘These schemes are related to the more general concepts of cognitive and systems modeling in the tradition of the semantic or conceptual structures of Osgood et al. (1975), the cognitive mapping of Axelrod (1976), and the structural modeling of Sowa (L984), Lastly, these analysis methodologies may be eonsidored only representative of potential algorithmic schemes that can impact future systems requirements analyses, Further study and application of these methods in operational and laboratory settings will provide an ensemble of techniques for motivating systems development and characterizing potential improvements in the data rich complex submarine environment, REFERENCES Atkin, R. and J. Casti (1977), Polyhedral Dynamics and the Geometry of Systems. International Institute for Applied Systems Analysis, Laxenburg, Austria Axelrod, R. (1976), “Structure of Decision,” In ‘The Cognitive Maps of Political Elites, Princeton University Press, Princeton, NJ. Barnes, R. M. (1980), Motion and Time Study: Design and Measurement of Work. John Wiley & Sons, New York, NY. Baum, A. and Y_M, Epstein (1975), Human Response to Crowding. John Wiley & Sons, New York. Chapanis, A. (1959), Human Engineering. Press, Baltimore, MD. Dean, L. M. W. M. Pugh, aad E. K. ©, Gunderson, (1975), "Spatial and Perceptual Comporents of Crowding Effects on Health and. Satisfaction,” Environment and Behavior, 7, 225-236. Research Techniques in ‘The Johns Hopkins 1191 Drillis, R. J. ( 1959), “The Use of Gliding Cyclograms in the Biomechanical Analysis of Movements,” Human Factors, April, 1-11 Expert Vision User's Manual (1986), Motion ‘Analysis Corporation, Santa Rosa, CA. Harary, F..R, Z., Norman and D. Cartwright (1865), Structural Models: Az Introduction to the Theory of Directed Graphs, John Wiley & Sons, New York, Hansen, L.T,, W, N, Hurwitz, and W. G. Madow (1953), Sample Survey Methods and Theory: Volume I. John Wiley & Sons, New York. Johnson, J. H.. (1988), Starpack User Manual. MeCain, G., V. C. Cox, and P. B. Paulus (1976), "The "Relationship Between fllness Complaints and Degree of Crowding in a Prison Environment,” Environment and Beltavior, 8, 283-290, Osgood, C. B,, G. J. Svie, and P. H. Tannenbaum (1975), "The Measurement of Meaning University of Ilinois Press, Chicago. Siegel, A. L, J. J. Wolf, and J. Pilitis (1982), "A "New Method for the Seientifie Layout of Worke spaces,” Applied Ergonomics, 13 (2), 67-90. Sowa, J. F. (1984), Conceptual Structures Information Processing in Mind and Machine, Addison-Wesley, Reading, MA. Stokols, D., M_Rall, B. Pinna, B., and J. Schopler (4973), "Physicsi, Social,” and Personal Determinants of the Perception of Crowding,” Environment and Behavior, 5, 87-115. "Submarine Command and Control Systems" (1983), NAVMATINST | 5450.27C, Department of Defense, Washington, DC, August. Tullis, T. S., B. Sperling, and F.H. Steinberg (1986), The use of multidimensional scaling for facilities layout; An application to the design of the space station. Proceedings of the Human Factors Society--30th Annual ‘Meeting. 38-42 Walker, W. E., J. M. Chaiken, and & J. Ignall (2975), Fire’ Department Deployment Analysis, a Public Policy Analysis Case Study, Elsevier North Holland, Inc, New Wallin, C. R. (1988), "Concept of Operations Experiment (COOPEX) 87-1 Summary Report,” TM 87-2122, Naval Underwater Systems Center, Newport, RI

You might also like