A Platoon-Level Model of Communication Flow and the Effects on Operator Performance

Transcription

1 A Platoon-Level Model of Communication Flow and te Effects on Operator Performance by Patricia W. Kilduff, Jennifer C. Swoboda, and Josua Katz ARL-MR-0656 November 2006 Approved for public release; distribution is unlimited.

2 NOTICES Disclaimers Te findings in tis report are not to be construed as an official Department of te Army position unless so designated by oter autorized documents. Citation of manufacturer s or trade names does not constitute an official endorsement or approval of te use tereof. DESTRUCTION NOTICE Destroy tis report wen it is no longer needed. Do not return it to te originator.

3 Army Researc Laboratory Aberdeen Proving Ground, MD ARL-MR-0656 November 2006 A Platoon-Level Model of Communication Flow and te Effects on Operator Performance Patricia W. Kilduff and Jennifer C. Swoboda Human Researc and Engineering Directorate, ARL Josua Katz MicroAnalysis and Design Approved for public release; distribution is unlimited.

4 REPORT DOCUMENTATION PAGE Form Approved OMB No Public reporting burden for tis collection of information is estimated to average 1 our per response, including te time for reviewing instructions, searcing existing data sources, gatering and maintaining te data needed, and completing and reviewing te collection information. Send comments regarding tis burden estimate or any oter aspect of tis collection of information, including suggestions for reducing te burden, to Department of Defense, Wasington Headquarters Services, Directorate for Information Operations and Reports ( ), 1215 Jefferson Davis Higway, Suite 1204, Arlington, VA Respondents sould be aware tat notwitstanding any oter provision of law, no person sall be subject to any penalty for failing to comply wit a collection of information if it does not display a currently valid OMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS. 1. REPORT DATE (DD-MM-YYYY) November TITLE AND SUBTITLE 2. REPORT TYPE Final 3. DATES COVERED (From - To) 5a. CONTRACT NUMBER A Platoon-Level Model of Communication Flow and te Effects on Operator Performance 6. AUTHOR(S) Patricia W. Kilduff; Jennifer C. Swoboda (bot of ARL); Josua Katz (MAAD) 5b. GRANT NUMBER 5c. PROGRAM ELEMENT NUMBER 5d. PROJECT NUMBER 62716AH70 5e. TASK NUMBER 5f. WORK UNIT NUMBER 7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) U.S. Army Researc Laboratory Human Researc and Engineering Directorate Aberdeen Proving Ground, MD SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT NUMBER ARL-MR SPONSOR/MONITOR'S ACRONYM(S) 11. SPONSOR/MONITOR'S REPORT NUMBER(S) 12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release; distribution is unlimited. 13. SUPPLEMENTARY NOTES 14. ABSTRACT Te Future Combat System (FCS) initiative is at te center of te Army s Objective Force Vision. Te Army Vision (2010) states tat U.S. forces must ave information superiority: te capability to collect, process, and disseminate an uninterrupted flow of information wile exploiting or denying an adversary's ability to do te same. In oter words, te Future Force will be empowered by dominant situational understanding. To predict ow proposed systems and displays will impact situational understanding and tereby decision making, te U.S. Army Researc Laboratory s Human Researc and Engineering Directorate used te tool C3TRACE (command, control, and communication: tecniques for reliable assessment of concept execution). C3TRACE is a modeling environment in wic one can develop multiple concept models for any size organization, staffed by any number of people, using any type of information tecnology, performing any number of functions and tasks, and under various communication and information loads. Among te performance measures tracked are Soldier utilization, te number of messages received, dropped, and interrupted, and te degree to wic te information is available to support Soldier decision making. C3TRACE was used to develop a platoon-level model of an FCS conceptual configuration in support of te situational understanding as an enabler for te unit of action maneuver team Soldiers Army Tecnology Objective. Te platoon model assumed wrist-mounted displays for te dismounted Soldiers and laptop-type displays for te mounted Soldiers. In general, te Soldiers using te laptop-type display were able to fully process more of teir incoming messages and made no decisions wit poor information quality. On te oter and, te majority of te Soldiers using te personal digital assistant ad ig utilization, iger numbers of dropped and interrupted messages, and decisions made wit poor information quality. Specifically, te squad leader and team leaders exibited te igest levels of utilization and decisions made wit poor information quality. Tis was not only because tey received more messages tan tose Soldiers in te mounted pase but also because of te tecnology being used. 15. SUBJECT TERMS decision making; information quality; utilization 16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT 18. NUMBER OF PAGES 19a. NAME OF RESPONSIBLE PERSON Patricia W. Kilduff a. REPORT Unclassified b. ABSTRACT Unclassified c. THIS PAGE Unclassified SAR 80 19b. TELEPHONE NUMBER (Include area code) Standard Form 298 (Rev. 8/98) Prescribed by ANSI Std. Z39.18 ii

5 Contents List of Figures List of Tables iv iv 1. Introduction 1 2. Model Development Organization and Personnel Equipment Assumptions Functions and Tasks Scenario and Communications Results Soldier Utilization Soldier Performance Decision Quality Discussion Conclusions and Recommendations Future Developments Summary References 17 Appendix A. Personnel Attributes 19 Appendix B. Task List for Eac Soldier 21 Appendix C. Scenario Description 57 Appendix D. Message Traffic From Scenario 59 Distribution List 71 iii

6 List of Figures Figure 1. Decay rate matrix....3 Figure 2. Infantry platoon personnel...4 Figure 3. Wrist-mounted PDA...5 Figure 4. PC wit touc screen capability....6 List of Tables Table 1. Example of message processing time for mounted Soldier (using laptop) versus dismounted Soldier (using wrist-mounted PDA)...6 Table 2. Pases of te scenario...8 Table 3. Utilization by Soldier...10 Table 4. Number of medium and ig priority messages received, interrupted, or dropped by igly utilized (45% or iger) Soldiers...11 Table 5. Number of dropped messages by pase for igly utilized operators...11 Table 6. Number of decisions based on information quality by Soldier iv

7 1. Introduction Te U.S. Army is in te midst of a revolutionary transformation and te Future Combat System (FCS) initiative is at te center. Te network-centric concept for te FCS is intended to empower te Soldier by providing accurate, timely, and actionable 1 information. Military missions will be conducted in an intensive information-ric, decision-making, and problem-solving environment, tus making information access and distribution critical to mission success. Tecnological advances promise a dramatic increase in te pace of information flow. At te same time, tis information-ric environment is certain to provide many callenges for te command, control and communication (C3) environment. Te potential for information overload, as distinct from information requirements, could negatively affect a Soldier s situation awareness (SA) (Endsley, 1995) and in turn, ave an adverse effect on is or er ability to make effective and timely decisions during a battle, ultimately diminising mission performance rater tan improving it. In order to improve te Soldier s ability to filter and select information, an improved understanding of information systems and te interaction of te systems wit te organizational structure must be acieved. A main objective of te Army tecnology objective (ATO) entitled Situational Understanding (SU) as an Enabler for te Unit of Action Maneuver Team Soldiers is to support and enance Soldier decision making in te proposed FCS information-saturated environment. Since te various concepts for te FCS organizational, personnel, and tecnological configurations cannot all be examined in uman-in-te-loop simulations or live exercises, constructive simulations and uman performance models are used. In particular, C3TRACE (command, control, and communication: tecniques for reliable assessment of concept execution) was used to model te communication flow and effects of information processing on Soldier and system performance. Te results elp to portray te limits in information processing and decision making of te Soldiers, wit te intent of affecting system design or implementation. C3TRACE (Kilduff, Swoboda, & Barnette, 2005) provides an environment for targeted evaluation of te effects of different configurations of Soldiers and information tecnology on performance: bot Soldier performance and overall system performance. Tis tool as an embedded discrete event simulation engine, Micro Saint Sarp 2 and a grapical user interface. It provides te capability to represent any organizational level, te people assigned to it, te tasks and functions tey will perform, and te communications patterns witin and outside te organization, all as a function of te frequency, criticality, and quality of incoming information. 1 Tat is, capable of being acted upon 2 Micro Saint Sarp is a trademark of Micro Analysis & Design. 1

8 A key C3TRACE metric is decision quality or te degree to wic te correct information is available to support a decision-making event. Te basis for te decision support metric is an embedded information-driven decision-making arcitecture (Wojciecowski, Wojcik, Arcer, & Dittman, 2001). Tis decision-making arcitecture uses 24 possible information elements taken from te U.S. Army s accelerated decision-making process documentation (Federation of American Scientists, 1996). Te information elements are grouped into six categories: (a) enemy force and actions, (b) friendly force and actions, (c) feasibility of te current plan, (d) suitability of te current plan, (e) acceptable risk, and (f) enemy course of action (COA) and potential COAs. C3TRACE uses tese elements in tree types of tasks: processing, collaborative, and decision making. Tese elements are attaced to te communication events tat trigger uman tasks. Information is collected in te processing tasks, sared in te collaboration tasks, and used in te decision-making tasks. For example, a processing task would be open/read message; a collaborative task would be discuss; and a decision-making task would be estimate impact. As te messages and te associated information elements move troug te task flow, C3TRACE accounts for wic Soldier knew wat elements of information and ow recent tat information may be to te Soldier. Te information elements are furter adjusted according to two factors: te accumulated information decay and te matc between available information and te information required to make a decision. As te model runs and messages are generated and sent to personnel to be read, te age of eac information element is generated, based on wen te assigned person first read te message and ow muc time elapsed before te person used te information in a decision task. It is a basic assumption of C3TRACE tat te older te information, te less useful it is. It is also assumed tat te rate of decay is not te same for all types of information (figure 1). Information about te location of te enemy is likely to cange frequently and rapidly, and tus, tat information element will decay quickly, wereas information about te mission objective of te friendly units is likely to remain stable over a given period of time. C3TRACE provides a decay algoritm to capture tis differential decay. Te matc between available information and te information required to make a decision occurs wenever a decision task is executed. Just as te information elements carried by eac message are identified, te information elements requirements of eac decision are identified. Does te decision maker ave te correct information, eiter processed directly from a message or received from a collaborator, to make a decision wen te time comes? In te end, te quality of a decision, tat is, te probability of making a good decision, is based on te matc between te information received by te decision maker and te information required to make a decision and also by ow muc te value of te information as decayed over time. Tis tecnique can elp to identify system and organizational inefficiencies, bottlenecks, or obstacles relevant to te ig quality and recent information required for effective decision making. 2

9 Figure 1. Decay rate matrix. For tis SU ATO-sponsored effort, a C3TRACE platoon-level model was built. Tis analysis examined te mounted and dismounted pases of a mission were te mounted Soldiers were assumed to be using laptop-type displays and te dismounted Soldiers were assumed to be using wrist-mounted personal digital assistants (PDAs). Te principal metrics of interest are (a) Soldier utilization, (b) te number and type of messages received, dropped, and interrupted, and (c) te degree of decision support. 2. Model Development Te development of te C3TRACE platoon-level model is described in te following subsections. 2.1 Organization and Personnel Te proposed FCS infantry company will encompass tree platoons and one company eadquarters (HQ) section and associated veicles (see figure 2). Eac platoon will consist of 49 personnel in five veicles (Unit of Action Maneuver Battle Lab [UAMBL], June 2003). In order to control te problem space and still capture te critical information flow, a subset of te total platoon was modeled. Te goal was to model te Soldiers wo would ave te most communication traffic and to simultaneously focus on te dismounted element. Te 16 personnel blocked off in figure 2 plus te 3

10 company commander (Co CDR) are te personnel modeled. Te Co CDR from te infantry company HQ section was added because of te significant communications traffic tat e would generate to te platoon leader (PL) and from te PL to te Co CDR. Tus, a total of 17 Soldiers was modeled and grouped into tree veicles according to te operational and organizational (O&O) plan (UAMBL, June 2003). Te Co CDR was in an infantry carrier veicle (ICV), operating outside te platoon s direct area of operation (AO). In reality, tere would be oter Soldiers in te veicle wit im but for te purposes of tis model, only te Co CDR was modeled. Te PL s veicle was an ICV and consisted of te PL, te veicle commander (VC), te driver (DRV1), a medic (MED), and a robotics noncommissioned officer (RNCO). Te PL was modeled as dismounted wile te remainder of te Soldiers in te veicle were modeled as mounted. Te platoon sergeant s (PSG) veicle was also an ICV and consisted of te PSG, a driver (DVR2), a squad leader (SL), two team leaders (TLA, TLB) and six infantrymen. Te SL, two TLs, and te six infantrymen were modeled as dismounted. For purposes of tis model and based on Future Force Warrior (FFW) guidance, te platoon model was conceptualized to use wrist-mounted PDAs for te dismounted Soldiers and laptop displays for te mounted Soldiers. Tese laptop displays were securely mounted in te veicle. Figure 2. Infantry platoon personnel. (Te blocked off personnel and veicles indicate te Soldiers included in te model, plus te Co CDR.) Eac of te Soldiers in tis platoon-level model was described in terms of is or er Soldier attributes or caracteristics. Te personnel caracteristics as modeled ere were used for 4

11 tracking purposes but do not directly affect te model execution or results. Te personnel caracteristics for eac modeled operator are sown in appendix A. For example, te PL was described as military occupational specialty (MOS)-11A, military training level-officer basic course, lengt of service-2 years, aptitude level-ig, age-22 to 25 years, rank-o2, time in position-4 to 6 monts, and igest civilian education-college. 2.2 Equipment Assumptions Altoug te Army is still developing te FCS and its family of equipment, te autors ad to make certain assumptions in order to proceed wit model development. Terefore, before building tis model, te autors made two assumptions about te equipment tat dismounted (see figure 3) and mounted Soldiers (see figure 4) would use, based on te FCS FFW equipment concept at tat time (tese devices are similar devices and do not represent any equipment undergoing development for FCS or FFW). Te first assumption was tat te dismounted Soldier would use a wrist-mounted PDA very similar to tat of a current PDA, wit a diagonal screen size of 3.5 inces. Te PDA, it is assumed, would sit on a Soldier s wrist orizontally in fasion unlike its commercial design of a vertical view. Furtermore, it is assumed tat te software running on tis Soldier s wrist-mounted PDA would support tis use. Te software running on te PDA is assumed to be a new developmental item tat provides a map, direction, and military symbology information. Figure 3. Wrist-mounted PDA. Te second equipment assumption was based on te mounted Soldier and te current FCS equipment at te time of model development. Te screen tat te mounted Soldier is assumed to interact wit is a tablet-like personal computer (PC) wit touc screen capability and approximately a 14.1-inc diagonal screen. It was tougt at te time of model development tat tis could offer portability to te mounted crew wen tey needed to dismount. Te veicle crew, wo is tougt not to dismount during operations, would ave a similar device mounted on te veicle (not removable). Te software assumed to be used in tis device will mirror capabilities similar to Force XXI Battle Command-Brigade and Below. 5

12 Figure 4. PC wit touc screen capability. C3TRACE was used to determine te performance effects of tese information display devices (laptop versus wrist-mounted PDA) on Soldiers riding in FCS veicles (mounted positions) and for Soldiers outside te FCS (dismounted). Most of te tasks required for manipulation of tese devices are similar. However, te wrist-mounted device requires additional time to process and access needed information. Tis additional time is illustrated in table 1 specifically open/read message, searc for digital information, find digital information, collaborate/discuss. Most of tese tasks are reading tasks tat require additional processing time. Te requirements of te wristmounted PDA may create decrements in performance tat could affect mission success. Table 1 provides an example of message-processing task times received troug by a mounted Soldier (RNCO) using a laptop display and by a dismounted Soldier (TLA) using a wrist-mounted PDA. Task times were provided by a subject matter expert (SME), wo was familiar wit te proposed FFW tecnologies; tese times served as input to te model. Table 1. Example of message processing time for mounted Soldier (using laptop) versus dismounted Soldier (using wrist-mounted PDA). Task Mounted Time (sec) Dismounted Time (sec) Receive Identify environment.5.5 Identify Sender Identify priority Pull up Open/Read Message Relevant to me.5.5 Collect Information Searc for Digital Information Find Digital Information Identify Actionable Coices Estimate Impact-Sort Collaborate - Discuss Finalize COA Take Action Update COA Average Time

13 2.3 Functions and Tasks According to Field Manual 7-8 (Headquarters, Department of te Army, 1992) and te FCS O&O (June 2003), te infantry platoon s mission is to close wit te enemy by means of fire, maneuver, and tactical assault to destroy, fix, or repel te enemy s assaults by fire, close combat, and counter-attack. Some specific infantry platoon mission-level battle drills include Platoon attack Squad attack React to contact Break contact React to ambus Knock out bunkers Enter building/clear room Enter/clear trenc Conduct initial breac of a mined wire obstacle Weter co-located or distributed, eac of te infantry tasks requires communication not only between veicles but also between personnel witin te veicles, between te mounted and dismounted personnel, between te dismounted temselves, and between ig ecelons, and oter platoons/companies. Te Co CDR uses is veicle to provide company command and control (C2) from distributed locations. In te model, te functions and tasks tat eac Soldier performed in processing communication events were based on guidance from a military SME wo was formerly an Army Ranger wit specific squad leader experience and currently works wit te FFW program (te tird autor on tis report). Te functions are composed of communication tasks consisting of te steps involved in processing eac message tat comes into te model. Wen one is considering te complete job of tese Soldiers, te communication is, of course, a subset of te mission. Te functions and tasks in te models were sequenced to reflect te process tat is expected to be followed under FCS missions. Eac task was assigned attributes including priority (ig, medium, low), SA level (perception, compreension, prediction), mode (manual, auto, bot), decision task (yes/no), collaborative task (yes/no), time to perform (distribution, mean, standard deviation), Soldier assignment (primary Soldier, alternate primary Soldier, supporting Soldier(s)), mental workload levels for Visual, Auditory, Cognitive and Psycomotor (VACP), and decision task information element weigting. A full listing of eac task in te model is sown in appendix B. Tis task listing also includes te mean time, standard deviation, primary operator, and VACP values. Witin te 17 model task flows (one per Soldier), tere were more tan 130 individual decision 7

14 tasks and included suc tasks as identify actionable coices, take action, or estimate impact. 2.4 Scenario and Communications Te SME-(tird autor)-developed scenario used for tis model was a 37-minute segment, based on an FFW related application tat begins 45 minutes into a mission in wic an infantry platoon moving in a mounted posture identifies and begins to clear a danger area. Tis particular segment was cosen because it ad a large volume of message traffic and would represent ig communication workload. Te scenario segment was divided into 23 pases, ranging from Pase 1 (identifying te danger area and planning were to stop) to Pase 23 (moving to assault at te bridge). It contains bot a mounted (Pases 1 troug 9) and dismounted segment (Pases 10 troug 23). Te complete list of pases and te lengt of te pase is sown in table 2. Te terrain in wic tis scenario is based is an open and rolling terrain environment. Tis scenario, including te times for eac pase, was provided by te SME and was derived from a FFW scenario. A full description of te scenario is provided in appendix C. Table 2. Pases of te scenario. Pase Description Time (in seconds) 1 Identify te danger area and planning were to stop Update mission profile to reflect danger area 48 3 Drivers told to stop 61 4 Move into security alt Launc unmanned aerial veicle (UAV) 83 6 Discussing planning for clearing te danger area SL briefs TL about wat and ow to manage te battle 71 8 UAV on station and provides feedback 44 9 PL reports to SL, TLs, and Co CDR 8 10 Initial movement of squad (dismounted Soldiers) Medic cecks status of Soldiers RNCO searcing danger area squad moving SL wants is squads to move into cover position SL is talking to RNCO and waiting for intelligence from UAV Handoff of sensor unmanned ground veicle (SUGV) from TLB to VC so 125 tey can figt instead of control robot 15 Moving to bridge SL and RNCO talking about wat UAV sees RNCO sends video to SL PL and RNCO entering fire missions based on video Squad stopping and pulling security SL explaining to TLs were to go next Stopped at last covered and concealed position PL talks to RNCO for update on wat e sees Squad stopped SL to RNCO intelligence up PL to RNCO to better position assets Moving to assault at bridge 128 Te communication events or messages map onto te scenario story and were developed by te SME. Eac message is defined by te type of communication (digital ( , witeboard, maps, sensors), voice, face to face, and written), te time it entered te model, tasks tat are triggered as a result of eac message, and te priority of te message. A complete list of te 8

15 messages entering te model is sown in appendix D. Te C3TRACE model works according to a basic input-trougput-output sceme. Tat is, te input to te task model is communication events, wic present an information event stream in a time sequence. As tese communication events enter te model, tasks are triggered and performed in a pattern tat reflects te logic for task brancing and interrupt priorities. Te logic pat (pat tat eac message followed troug te organization) for eac message was determined on te basis of message content and its timing in te scenario. 3. Results Te focus of te model was on information flow and communication, particularly wen te information flow led to a decision. Model results included measures of performance suc as Soldier use, Soldier performance in terms of received versus interrupted and dropped messages, and decision quality as a function of information goodness and matc to te decision requirements. 3.1 Soldier Utilization Table 3 sows te utilization for eac Soldier over te duration of te scenario. Utilization is defined as te percentage of time tat a particular Soldier was busy during te scenario. For example, te scenario for tis model ran for 2,250 seconds or 37 minutes. A Soldier s utilization is te percent of tat 37 minutes tat s/e is busy receiving, sending, and tinking about te content of eac message and te resulting action required. In oter words, during te 37-minute scenario, te SL was busy for 86% of tat time. Utilization rates are for communication tasks only and do not include oter mission-related tasks tat are not included in te model. Te Soldiers are grouped by mounted (7 Soldiers) and dismounted (10 Soldiers) according to teir position in te platoon. Te dismounted Soldiers were modeled solely performing dismounted tasks and te mounted Soldiers performing mounted tasks. For te mounted Soldiers, only te PSG and RNCO ad utilization rates above or approacing 50%. On te oter and, for te dismounted Soldiers, all utilization rates were close to 50% or iger, but squad and team leaders ad te igest utilization rates. Te autors considered utilization at or above 45% to be ig. 9

16 Table 3. Utilization by Soldier. Soldier Utilization (percent) Mounted Co Cdr 4 VC 23 DVR1 11 MED 22 PSG 49 RNCO 56 DVR2 8 Dismounted PL 71 SL 86 TLA 77 Rifleman 1A 45 Rifleman 2A 54 Rifleman 3A 49 TLB 83 Rifleman 1B 61 Rifleman 2B 62 Rifleman 3B Soldier Performance Table 4 sows te Soldier performance data in terms of te total number of messages received, number and percent of messages dropped, and number of times messages were interrupted during te scenario. Te data in tis table represent te Soldiers wo were utilized 45% of te time and iger. Altoug most of te Soldiers ad utilization exceeding 50%, several Soldiers were very close to tis level (45% or >). Actually, tis included all (100%) of te dismounted Soldiers and two (20%) of te mounted Soldiers. Te SL received a total of 124 messages and ad 286 task interrupts and 88 message drops. A task interrupt is defined as a message for wic te processing is interrupted because anoter message of a iger priority comes into te model causing te Soldier to stop working on te current message (te message is suspended at tis point) and start working on te new one. If te Soldier is not able to return to te interrupted message to finis processing it witin a predefined 10-minute time limit, tat message is dropped and not processed any furter. It is important to note tat if a message is interrupted and consequently dropped, it may lead to a critical piece of information being lost, tereby aving a potentially significant effect on te outcome of te battle. Te same could be said for a message tat as been interrupted several times and ten ultimately processed. In tis particular case, te information would most likely be old. If a message was dropped during a collaboration task or a decision-making task, tis, too, could ave a significant effect on te battle outcome. A message can be interrupted more tan one time, wic was wy, in some cases, te number of interrupts was iger tan te number of messages received. Among te mounted Soldiers wit utilization less tan 45%, few messages were dropped. Among te dismounted Soldiers, te squad and team leaders ad te igest number (and percentage of) message drops. 10

17 Table 4. Number of medium and ig priority messages received, interrupted, or dropped by igly utilized (45% or iger) Soldiers. Soldier Number of Messages Received Number of Dropped Messages Percent Dropped Messages Number of Times Messages Interrupted Mounted PSG RNCO Dismounted PL SL TLA Rifleman 1A Rifleman 2A Rifleman 3A TLB Rifleman 1B Rifleman 2B Rifleman 3B It is interesting to identify during wic pases of te scenario te majority of messages were dropped. Table 5 sows te number of messages dropped per pase. For most of te pases until Pase 11, tere are few drops. Tere are 10 message drops in Pase 4 because te veicle is getting ready to stop so te Soldiers can dismount. During Pases 1 troug 9, all Soldiers are mounted. Pases 12 troug 23 are were te bulk of te mission-related action takes place. Dismounts begin to assume teir position on te ground and take action. During tese later pases, te majority of messages were dropped by te dismounted Soldiers. Table 5. Number of dropped messages by pase for igly utilized operators. Pase Number of Drops 1 Identify te danger area and planning were to stop 0 2 Update mission profile to reflect danger area 0 3 Drivers told to stop 4 4 Move into security alt 10 5 Deploy UAV 4 6 Discussing planning for clearing te danger area 3 7 SL briefs TLs about wat and ow to manage te battle 7 8 UAV on station and provides feedback 1 9 PL reports to SL, TLs, and Co 1 10 Initial movement of squad (dismounted) MED cecks status of soldiers RNCO searcing danger area squad moving SL wants is squads to move into cover position SL is talking to RNCO and waiting for intelligence from UAV Handoff of SUGV from TLB to VC so tey can figt instead of control robot Moving to bridge SL and RNCO talking about wat UAV sees RNCO sends video to SL 4 17 PL and RNCO entering fire missions based on video Squad stopping and pulling security SL explaining to TL s were to go next Stopped at last covered and concealed position PL talks to RNCO for update on wat e sees 8 21 Squad stopped 5 22 SL to RNCO intelligence to PL to RNCO to better position assets Moving to assault at bridge 19 11

18 Dismounts must rely on teir wrist-mounted PDA for any communication weter from iger or from oter dismounts during tis period of ig activity. 3.3 Decision Quality Table 6 sows te number of decisions made wit good or bad quality information for te mounted Soldiers, including te PSG, and RNCO, and te dismounted Soldiers, including te PL, SL, TLA, rifleman 1A, 2A, 3A, TLB, rifleman 1B, 2B, and 3B. It also sows tat most decisions were made wit good quality information except for te squad leader. Decisions made wit good information quality are tose in wic a Soldier ad te most recent information available and was able to process and act upon tat information in a timely manner. However, decisions made wit poor information quality are tose in wic a Soldier ad only old (out of date) information, possibly as a result of competing tasks and interruptions and was not able to act in a timely manner in making a decision. Te assumption in C3TRACE is tat decisions are only as good as te information used to make tem. Te quality of te information and te degree of te matc of te information available to te information required equals te quality of te decision. If a Soldier is provided wit poor information, te likeliood of making a poor decision is ig. For te purpose of tis model, te probability of making a good decision is defined as any decision wit 50% good quality information, and conversely, te probability of making a bad decision is defined as any decision wit <50% good quality information. Te rationale beind tis fairly liberal cutoff is tat decisions made wit poor information quality do not always lead to a negative tactical outcome. Individual personnel caracteristics suc as experience, training, leadersip qualities, and so fort, can all ave an impact on decision making as well as recency of te information. Table 6. Number of decisions based on information quality by Soldier. Soldier Number of Decisions Made wit Good Info Quality Percent of Decisions Made wit Good Info Quality Number of Decisions Made wit Poor Info Quality Percent of Decisions Made wit Poor Info Quality Total Number of Decisions Mounted RNCO PSG Dismounted PL SL TLA Rifleman 1A Rifleman 2A Rifleman 3A TLB Rifleman 1B Rifleman 2B Rifleman 3B

19 4. Discussion As mentioned earlier in te report, a subset of te total platoon was modeled. Tis subset included a total of 17 Soldiers, essentially one squad, of 49 Soldiers (50 wit te Co Cdr). Te results tat are presented and discussed are relevant to only te subset. In reality, tere would be four squads. As a result of te increased number of personnel tat would comprise te oter tree squads, te Co Cdr s message traffic would increase because of communications from te additional personnel. Te SL s message traffic would also increase because of message traffic between SLs. Te Soldiers igligted in te results were te PSG, te RNCO, te PL, te SL, TLA, TLB, riflemen 1A, 2A, 3A, 1B, 2B, and 3B. Of tose 12 Soldiers, all were dismounted except te PSG and te RNCO. As mentioned earlier, te dismounted personnel used a wrist-mounted display wile te mounted personnel used a laptop display. Te PL received messages from te Co CDR (iger), from te SL (lower), and from display update messages tat are event or system generated. As a result of te volume of messages tat e was receiving, e ad a ig number of interrupts and was 71% utilized. Even toug e ad a ig number of interrupts, e ad only tree dropped messages, wic indicated tat e was able to finis processing te most of is received messages, including te interrupted ones. Te tree messages tat were dropped were all ig priority. Te SL also received messages from iger (PL) and lower (TLs) and ad te igest number of interrupts and drops of any of te Soldiers. More tan 50% of te SL s decisions were based on poor information quality. Te presumed low quality decisions for te SL may ave been because e was eavily utilized at 86%, ad too many messages to respond or attend to, ad additional messages to send to is TLs based on wat e was receiving from iger eadquarters, and te additional responsibility for te personnel wit im in a dismounted status. TLA received messages from iger (SL), TLB, and lower (A team riflemen) during te course of te scenario. He ad a ig number of interrupts and a ig number of drops as well. Tis is all wile leading A Team in a battle on te ground. Te RNCO was not only responding to messages from upper and lower personnel but also dealing wit sensor information. Tis is in addition to is duties in dealing wit robot-controlrelated messages. Tis model dealt exclusively wit communication-related tasks and did not model te RNCO robotic control tasks. Because te non-communication tasks workload was not modeled, te RNCO did not ave a ig number of interrupts or drops (3) wic meant e was able to get is messages fully processed. Since e was able to do tis in a timely manner, e did not make any decisions wit poor information quality. Note tat tis particular scenario did not eavily involve te RNCO but focused instead on te dismounts. 13

20 Te riflemen were only receiving messages from iger eadquarters as well as generating messages to iger eadquarters. Individually, tey received between 79 and 82 messages and were utilized between 45% and 64% over te lengt of te scenario. In addition to dealing wit messages tey were also dealing wit ground operations and related tasks. Tis is evidence tat te dismounts are required to multi-task because tey are moving on te ground and responding to communications simultaneously. Te majority of Soldiers ad good quality information tat led to an optimum decision. Te SL was te only one wo ad a large number of decisions made wit poor information quality (>50%). Te fact tat tis Soldier was interrupted wen processing a message, did not receive updated information in a timely manner, or was over utilized to te point of missing information could all contribute to a poor quality decision. In summary, of te 12 operators focused on in te Results and Discussion sections, two used te veicle-mounted laptop display wile te oter 10 used te wrist-mounted PDA display. In general, te operators using te laptop display were able to fully process more of teir incoming messages and made no decisions wit poor information quality. Te majority of te operators using te PDA ad ig utilization, iger numbers of dropped and interrupted messages, and decisions made wit poor information quality. 5. Conclusions and Recommendations Tis model was designed to investigate operator performance wen a laptop display was used in a mounted setting, and a wrist-mounted PDA was used in a dismounted setting. Altoug not intentional, te model also provided uman factors (HF) information tat indicated some of te reasons wy te wrist-mounted PDA migt not be te best device for use by te dismounts as tey process communication messages. Te insigt provided ere is based on te results of te modeled use and manipulation of te equipment and te scenario. Te results of tis model and scenario sowed poor performance by te majority of te dismounted element because tey were utilized at a ig level. Tese Soldiers ad too many messages to process during dismounted pases and made occasional decisions wit poor information quality. Te information quality was poor because te information was old. As a result of aving too many messages to process, te Soldiers were not able to process tem in a timely manner, wic resulted in decisions based on poor information quality and messages tat were dropped before processing could be completed. It is important to note tat tis scenario lasted only 37 minutes and involved only planning and maneuver communications events. Tere was no engagement by te enemy in tis scenario. In suc a case, it is likely tat te number of decisions wit poor information quality and dropped and interrupted messages would increase. 14

21 In te model, it was assumed tat all conditions were perfect. Tis includes initial information quality at 100%, perfect weater, perfect ligting conditions, no nigt exercises, no missionoriented protective posture (MOPP), or cold-weater gear, etc. Some reasons as to wy te PDA would not be te optimal tecnology for te dismounts are (a) te dismounted operators were multitasking by concentrating on teir maneuvers wile responding to incoming messages, (b) te display area and te individual keys of te keyboard on te PDA were muc smaller tan wit te laptop display, resulting in increased reading and response time, i.e., reference table 1. Tere was not sufficient time for te manipulation required of wrist-mounted PDA in a dismounted environment. Manipulation of te device included scrolling troug messages and use of te stylus to enter and searc for desired information. Altoug not considered as part of tis modeling effort, te keyboard, stylus, and te manipulation of te screen wen te user was scrolling could be affected by te use of gloves, dayligt or nigt conditions. Too muc information on a very small screen and weater can also ave an adverse effect. Te dismounted personnel sould not be taxed to look at a screen tat could be difficult to read. Teir primary concern sould be keeping temselves and oter Soldiers alive. As mentioned earlier, dismounts, by te very nature of teir responsibilities, are multi-tasking. As suc, tey sould be provided wit equipment tat is simple and easy to use and requires te least manipulation from tem. Tey sould be provided wit a reliable radio system and assistive tecnologies tat are ands free (for example, a aptic belt). Based on te scenario, te dismounts ad too muc communication traffic, not always to respond to, but to be aware of. Even tese types of communication events sometimes require action. Based on te volume of messages directed to te dismounts, it is impossible for te Soldiers to attend to eac one. It would be beneficial for a filtering system to be employed so tat only important and necessary information was forwarded to a given dismounted Soldier. Also, limiting te amount of incoming information troug filtering sould be considered. Te results of tis platoon model configuration, including personnel, equipment, and scenario, indicate tat tose dismounted Soldiers wearing a wrist-mounted display exibited ig levels of utilization and made decisions based on poor information quality. Overall, te (mounted) Soldiers using laptops exibited low levels of utilization and made all decisions based on good information quality. It is reasonable to assume tat te difference in operator performance is attributable to te seer volume of messages received by mounted Soldiers compared to dismounted Soldiers and te equipment tat is being used. It is terefore recommended tat a wrist-mounted PDA be designed for ease of reading, rapid processing times, and low manipulation requirements. It would also be feasible to explore alternate tecnologies requiring less manipulation for use by te dismounted personnel. 15

22 6. Future Developments Ensuing work includes modeling te dismounted Soldiers wit an alternate display design suc as a aptic belt. 7. Summary Te use of C3TRACE, in wic te capability to easily represent any ecelon level, te people assigned to tat organization, te tasks and functions tey will perform, te equipment tey will use, and a communications pattern witin and outside te organization, will allow for wat-if evaluations of numerous concepts witout te need for live exercises or experiments. Te model in tis report represented platoon-level personnel wit wrist-mounted PDA tecnology use for dismounts. Model results sowed tat tis tecnology ad ig utilization levels associated wit it. Furter exploration of alternate tecnologies is recommended. Because tese results were based on a simulation model, te C3TRACE capability saved time and money as well as provided an evaluation of a critical information tecnology, wic if accomplised by umanin-te-loop experiments would be extremely time consuming and expensive. Te SU ATO will transition display guidelines to te FFW and FCS programs to enance te SU of Soldiers in te UA. Ensuing models will be developed to explore te use of alternate tecnologies tat will enance Soldier SU. 16

23 8. References Federation of American Scientists. Military Intelligence Officer Advanced Course. Offensive Operations and Intelligence (TCAICE): Tactical Decision Making Process [on line], Web site: ttp:// Department of te Army. Vision 2010 [on line]. Web site: ttp:// Endsley, M. R. Measurement of Situation Awareness in Dynamic Systems. Human Factors 1995, 37(1), Headquarters, Department of te Army. Infantry Rifle Platoon and Squad; Field Manual 7-8; Wasington, DC, Kilduff, P. W.; Swoboda, J. C.; Barnette, D. B. Command, Control, and Communications Tecniques for te Reliable Assessment of Concept Execution (C3TRACE) Modeling Environment: Te Tool; ARL-MR-061); U.S. Army Researc Laboratory: Aberdeen Proving Ground, MD, Unit of Action Maneuver Battle Laboratory. Te United States Army Objective Force Operational and Organizational Plan Maneuver Unit of Action. Training and Doctrine Command (TRADOC) Pamplet O&O (final), Fort Knox, KY, 30 Jun Unit of Action Maneuver Battle Laboratory. Operational Requirements Document for te Future Combat System. Fort Knox, KY, Wojciecowski, J. Q.; Wojcik, T.; Arcer, S.; Dittman, S. Information- Driven Decision Making Human Performance Modeling. Proceedings of te 2001 Military, Government and Aerospace Simulation Symposium: Advanced Simulation Tecnologies Conference. Seattle, WA,

24 INTENTIONALLY LEFT BLANK. 18

25 Appendix A. Personnel Attributes Veicle Soldier Name MOS Military Training level Lengt of Service Aptitude level Age Rank Time in position Higest civilian education Maximum No. of Lead Tasks ICV 1 PL 11A IOBC* 2yrs ig O2 4-6 monts College 3 Medic 91B AIT* 2yrs ig E4 1-2 years RNCO 11B PLDC* 3 years ig E years DRV 11B AIT* 2yrs medium E monts Veicle Commander 11B PLDC* 3 years medium E monts Hig scool 3 Hig scool 3 Hig scool 2 Hig scool 1 ICV 2 PSG 11B ANCOC* 6-10 years ig E monts SL 11B BNCOC* 5 years medium E years A Team Leader 11B PLDC* 3yrs ig E monts Rifleman 1A 11B AIT* 2 yrs medium E-2 3 monts Grenadier 2A 11B AIT* 3yrs medium E-3 Automatic Rifleman 3A 11B AIT* 2yrs medium E monts monts B Team Leader 11B PLDC* 3yrs ig E years Rifleman 1B 11B AIT* 2 yrs medium E-2 Grenadier 2B 11B AIT* 3yrs medium E-3 Automatic Rifleman 3B 11B AIT* 2 yrs medium E monts monts monts monts Hig scool Hig scool Hig scool Hig scool Hig scool Hig scool Hig scool Hig scool Hig scool Hig scool Hig scool DRV 11B AIT* 3yrs medium E CO HQ CO 11A CAS3* years ig O3 7-9 monts College *(IOBC-Infantry Officer Basic Course, AIT-Advanced Individual Training, PLDC-Primary Leadersip Development Course, ANCOC-Advanced Non-Commissioned Officer Course, BNCOC-Basic Non-Commissioned Officer Course, CAS3-Combined Arms and Services Staff Scool

26 INTENTIONALLY LEFT BLANK. 20

27 Appendix B. Task List for Eac Soldier Task Name Time (sec) SD (sec) Veicle Operator Visual Auditory Cognitive Perceptual Team Leader (TLA) START RECEIVE MSG/UPDT 0 0 ICV 2 TLA COLLABORATE 0 0 ICV 2 TLA Listen ICV 2 TLA No action 0 0 ICV 2 TLA Pause 1 30 ICV 2 TLA No action 0 0 ICV 2 TLA PL Listening ICV1 PL MED Listening ICV1 Medic VC Listening ICV1 VC CO Listening Co HQ CO SL Listening ICV 2 SL 0 0 Driver1 Listening ICV1 DVR Radio out for info (RFI) ICV 2 TLA Driver2 Listening ICV 2 DVR PSG Listening ICV 2 PSG TLA Listening ICV 2 TLA TLB Listening ICV 2 TLB Rifleman1A Listening ICV 2 Rifleman 1A Rifleman2A Listening ICV 2 Grenadier 2A Rifleman3A Listening ICV 2 Auto Rif 3A Rifleman1B Listening ICV 2 Rifleman 1B Rifleman2B Listening ICV 2 Grenadier 2B Rifleman3B Listening ICV 2 Auto Rif 3B COLLECT INFO ICV 2 Branc 0 0 TLA Listen ICV 2 TLA Relevant to me ICV 2 TLA Not relevant to me ICV 2 TLA ESTIMATE IMPACT-- ICV 2 Sort TLA Finalize COA--Sort 10 5 ICV 2 TLA Update mission profile-- ICV 2 Sort TLA Relevant to me ICV 2 TLA Searc for Digital Info ICV 2 TLA Not relevant to me ICV 2 TLA Relevant to me ICV 2 TLA Not relevant to me ICV 2 TLA Syncronize wit ICV 2 Higer Database--Sort TLA Haptic Message 0 0 ICV 2 TLA Recognize ICV 2 TLA ID Message ICV 2 TLA Direction 0 0 ICV 2 TLA Reply to voice ICV 2 TLA Alert 0 0 ICV 2 TLA Warning 0 0 ICV 2 TLA Caution 0 0 ICV 2 TLA ID Direction 0 0 ICV 2 TLA

28 ID Alert 0 0 ICV 2 TLA ID Warning 0 0 ICV 2 TLA ID Caution 0 0 ICV 2 TLA DISCUSS (TL A) ICV 2 TLA Cange recognized 2 2 ICV 2 TLA ICV 2 TLA Syncronize wit ICV 2 Higer Database--Long TLA Receive radio info ICV 2 TLA ESTIMATE IMPACT-- ICV 2 Long TLA Discuss (PL) ICV1 PL Pull up ICV 2 TLA Internal Trigger 0 0 ICV 2 TLA ID Current Mission ICV 2 TLA ID Environmental ICV 2 Canges TLA ID environment ICV 2 TLA ID wo from ICV 2 TLA ID priority ICV 2 TLA Find digital info ICV 2 TLA TAKE ACTION 0 0 ICV 2 TLA Voice Msg 0 0 ICV 2 TLA Verbal request ICV 2 TLA Receive verbal ICV 2 response TLA Voice Out ICV 2 TLA Send request ICV 2 TLA Receive response ICV 2 TLA START MONITOR 0 0 ICV 2 TLA See Visual Cange ICV 2 TLA Display update 0 0 ICV 2 TLA Pause 1 30 ICV 2 TLA ID ACTIONABLE ICV 2 CHOICES TLA START LISTEN 0 0 ICV 2 TLA Mission Profile Update ICV 2 TLA Discuss (TL B) ICV 2 TLB Open/Read Message ICV 2 TLA Discuss (Rifleman A) ICV 2 Rifleman 1A Discuss (Grenadier A) ICV 2 Grenadier 2A Discuss (Automatic ICV 2 Rifleman) Auto Rif 3A Finalize COA--Long ICV 2 TLA Update COA ICV 2 TLA Perform pysical action ICV 2 TLA Update mission profile-- ICV 2 Long TLA ID functionality ICV 2 TLA Switc to appropriate ICV 2 I/F TLA ID cannel ICV 2 TLA Switc to appropriate ICV 2 cannel TLA