Exercise Phoenix 98 to support Army development of operating procedures for an Armed Reconnaissance. Helicopter (ARH) in conjunction with the ARH ...
Concept Demonstrator Situation Awareness System for the Australian Army R. S. Seymour, AM. Grisogono and J. D. Krieg Defence Science and Technology Organisation Land Operations Division PO Box 1500 SALISBURY SA 5108
Keywords: Situation Awareness, Synthetic Environment, Land, Army, Information Management ABSTRACT: A concept demonstrator Situation Awareness system is being developed for the Australian Army. The intent is to illustrate possibilities in battle-space visualisation, decision support, information dissemination and information management relevant to the Restructuring of the Army (RTA) process. The demonstration system being assembled will incorporate advanced GIS capability for situation display, Intelligent Agent technology applied to information fusion and surveillance asset management, capability for collaborative mission planning, mission rehearsal and real-time mission monitoring. The GIS system will provide detailed 3-D terrain visualisation and interactive 'flythrough’ capability. It will show deployment of own forces and in particular surveillance system coverage using realistic sensor models with environmental factors included. Enemy and other detections by the surveillance system will be displayed preparatory to information fusion. The fusion will be guided by Intelligent Agent techniques, which are structured around the Intelligence Preparation of the Battlefield (IPB) process, which aims to produce predictions of enemy courses of action (COAs). Intelligent Agents will also guide the management of the surveillance system to collect further information to assist in refinement of the enemy COAs. Own force response planning is the next step in the Military Appreciation Process (MAP) and simulation techniques to assist this will be incorporated. Consistency of awareness across dispersed locations and timely access to, and dissemination of, information, by all levels of decision makers, are priority considerations. The effort is a collaboration of several private contractors with DSTO. An initial demonstration of the system occurred at Exercise Phoenix in September 1998 in conjunction with the virtual Armed Reconnaissance Helicopter simulation, and further demonstrations are planned. 1.
Introduction
Situation Awareness can be defined as “the appreciation by the decision maker of the sum total of all the information necessary to make an optimal decision”. Land Operations Division, DSTO, has been developing concepts in Situation Awareness (SA) for the Australian Army over the past year and a half and has funded several contracts with private industry to develop initial concept demonstrations. These developments covered a wide range of computer based situation visualisation and decision support tools, but were intentionally not particularly focused on a specific Army task. An opportunity arose to field a demonstration at Exercise Phoenix 98 to support Army development of operating procedures for an Armed Reconnaissance Helicopter (ARH) in conjunction with the ARH mission simulation described in the paper by Grisogono et al (this conference, Synthetic Environments in Support of Capability Development: an Armed Reconnaissance Helicopter Case Study at Exercise Phoenix). Of prime interest was to optimally task and utilise the asset to provide rapid, long-range target acquisition and engagement. With the potential of hostile targets to possess significant firepower (eg SAMs), the vulnerability of the ARH was an important consideration in developing suitable tactics. Shared Situation Awareness, if it can be made available, can have a dramatic influence on such tactics.
A Situation Awareness system has the potential to greatly enhance the operational capability of an ARH by minimising time delays in decision making. It is also feasible for the SA system to reduce the target search area by employing aids to predict enemy courses of action (COAs). The SA system might also assist mission planning and rehearsal by providing information on potential enemy capabilities and by enabling 3D visualisation of the target area. As no existing asset was deemed close enough in capability to an ARH, it was decided to develop a computer based simulation. This simulation, termed the virtual ARH mission simulator (ARH Sim), will be described briefly below and in detail in other papers at this conference. The SA concept demonstration system was integrated with the ARH Sim both physically and in the trial procedures for tasking and conducting the mission. This report will detail how some important concepts in SA were implemented and demonstrated for the ARH specific case. The aims of the demonstration were: • To raise user awareness of the new concepts and the potential of SA tools. Indirectly this heightened user awareness would potentially influence acquisition projects such as the Battlefield Command Support System (BCSS) through the Joint Application Development process that Project is employing. • To aid the Task Force in developing operating procedures for tasking and conducting missions, in the context of the computer based decision support
tools they could reasonably expect to be provided by projects such as BCSS. • To establish the utility of fielding prototype or concept demonstrator systems as part of a synthetic environment immersed in a real Army field exercise. • To obtain feedback form users to set priorities for further R&D in SA. This report will describe how all these objectives were met. 2.
Some Concepts in Situation Awareness
The definition of Situation Awareness given at the start of the Introduction section of this paper, is in accord with that adopted by the Commander Australian Theatre (COMAST). It brings out the importance of the human decision maker, and leads to an understanding of ‘optimal decision’ as implying a trade-off between timely but good, and ideal but too late. It also shows that SA is part of the bigger C4ISR system encompassing everything from information collection through to decision making and its impact on the battle. In particular, “sum total of all the information necessary”, implies information fusion, the tailoring of the information to meet the needs of the decision maker, and that different decision makers in the command hierarchy will have different needs. Fundamental to situation visualisation in the Land environment is representation of terrain. Terrain and its vegetation can provide cover for both friendly and enemy forces and will constrain mobility of both. Additionally, most information (but not all) is associated with a specific location, and it is natural and convenient for users to access this information via its location on a spatial display (ie a map) of the region of interest. State of the art computer technology can enable realistic visualisation of terrain in 3D where a user can ‘fly-through’ and examine things from different perspectives. The spatial relationship of entities on the terrain (friendly and enemy dispositions) can be quickly appreciated, and missions planned and rehearsed in the virtual environment. A variety of other information can be keyed to the spatial display by providing links to databases via ‘point and click’ on the maps. Such information might include: still imagery; video clips; other sensor data; text reports; audio reports; etc. Armies have always used maps for battlefield visualisation but modern computer technology offers considerable possibilities for enhancement. Some of these possibilities were the basis of the Phoenix SA demonstration. Multiple users in the Command hierarchy have different tasks and different information needs. However all the tasks need to be linked and directed toward a common goal. There needs to be collaboration in decision making and planning. All this prescribes information availability and sharing in a timely and consistent fashion. A concept that has emerged recently is that of the ‘virtual information space’, where all users have access to a consistent set of information but can each be putting it to a different use. If one user adds to the pool, this is available in near real time to all other
users. With this concept, users of quite different applications would share the same data and would thus be interoperable. At Phoenix the SA demonstration used an implementation of the virtual information space to provide several different users access to a timely, consistent pool of information. We were able to demonstrate at Phoenix: • consistent depictions of enemy and own force dispositions across all users; • collaborative planning between remote locations; • mission rehearsal; • mission monitoring; • in cockpit SA and; • the ability to re-task mid mission. 3.
The Virtual Armed Reconnaissance Helicopter Mission Simulator (ARH Sim)
This section will give an abbreviated description of the synthetic environment that was established at Exercise Phoenix sufficient only to see how the SA demonstrator integrated with the virtual part of the environment. Full detail of the virtual part will be given in other papers at this conference. The ARH Sim consisted of a virtual reality system where pilots and crew could navigate a computer model of a helicopter around a virtual terrain to locate virtual targets inserted in the terrain. Two helicopters were simulated and linked such that each was aware of the other. The ARH Sim comprised several components: • Two Helicopter cockpits with two crew in each. The pilot had controls to ‘fly’ the helicopter and the gunner (commander) had weapons control and the ability to control a look direction (which could be considered as an imaging sensor). Crew could communicate via voice over radio with each other, the other helicopter and with the Multi-Role Aviation Battalion Headquarters (MRAB-HQ). • The Boeing Advanced Tactical Combat Model (ATCOM) for simulation of helicopter dynamics (how the helicopter reacted to the flight controls), which calculated the helicopter position and attitude. ATCOM also provided a multi-function display for the pilots’ use and a simulation of a radar sensor. This simulation appeared to incorporate a line-of-sight calculation and was able to detect virtual targets in the virtual terrain from stand-off distances of several kilometres. • The ModSAF simulation, which was intended to be the core module in the ARH Sim. It was linked to all the others via DIS protocol. ModSAF was to maintain the overview of all entities and their relationships in the simulation but modification to this intent was necessary as described in another paper at this conference (Menadue, W.I. and Robertson, S.W.H., Implementation of a Low-Cost Virtual Helicopter for Exercise Phoenix).
•
•
problem with AUSTACSS. Extractions were the positions of the two virtual helicopters, and these were supplied to the SA systems at an update interval of about 20 secs (to simulate the relaying of GPS locations from real helicopters). DICE communicated with the real world in appropriate message formats. ADFORMS were used to communicate with AUSTACSS and a customised format was developed to communicate with the ‘virtual information space’ of the SA system (see later). It was also possible to manually insert entities into DICE, which could then be communicated to both the virtual and SA systems. A schematic of the ARH Sim is shown in Fig. 1.
Stealth-views to generate out-the-window views for each helicopter, with extensive terrain models comprised of digital terrain elevation data (DTED) draped with imagery (described in detail in another paper here, Perry, A.R., Robertson S.W.H., Fowler G.L. and Menadue W.I., Virtual Terrain for a Helicopter Mission Simulator). DICE (Distributed Interactive C3I Effectiveness) [1], which was used to provide the two-way interface between the virtual and real worlds. Through its DIS interface it was able to insert real world entities and extract positions of virtual entities. Insertions were intended to be entities obtained from the Australian Army Command Support System (AUSTACSS, the precursor to BCSS), however this did not eventuate in the field due to a to real world
DICE
Stealth 1
ModSAF
Stealth 1a
ATCOM 1
Stealth 2
Stealth 2a
ATCOM 2
Helo 1
Helo 2
Figure 1. The Helicopter Mission Simulator, (ARH Sim).
4.
The Fielded Situation Awareness System
Two major systems were fielded at Phoenix to provide the SA functionality relevant to the tasking and conduct of the ARH missions. These systems, provided by the companies Maptek and Exa-min, were developed under contract to DSTO but each system was based on a considerable amount of proprietary development. LOD produced the interfacing to both systems and to DICE to provide the concept demonstration of ‘the virtual information space’. All systems ran on PCs with advanced graphics capability under the Windows NT environment and together with the MS Office 97 suite. In particular MS Access was used to provide storage of ‘change’ information necessary to update each individual instance of a system in case of network failure or system crash. Further description of some of the components is given in the following sub-sections 4. 1 The Network There were five SA workstations configured in a LAN and physically connected by multimode optical fibre links. One station at Task Force Headquarters (TFHQ) was some kilometres distance from the others, which were all grouped around the Multi-Role Aviation Battalion (MRAB). Two stations were in the MRAB-
HQ and one in each of the ARH cockpits, to the left of the Commanders multi-function display. All stations were configured identically with the full suite of software and each was preloaded with the terrain data for the ‘play-box’ area east of Tindal. The hardware configuration was either: CPU dual 300MHz pentium or single 350 MHz pentium HDD 4 Gbyte RAM 128 Mbyte Graphics Diamond FireGL 4000 w/32MB or
Intel i740 AGP w/8MB Network
100BaseT4 Ethernet
4. 2 The Maptek System The Maptek system was based on their ‘Carmen’ architecture with special development tailored for the Phoenix demonstration. Carmen enables display of spatial (geographic) data in either 2 or 3D, with virtual ‘fly through’ capability in the 3D mode. The customised features included: • display of 3D models for the helicopters;
•
Examples of the ‘push mode’ transfer were: • The user entered a new threat zone on one of the workstations using the Maptek software, which was then pushed out to the other stations, • The DICE connection to the virtual world pushing in updated information about the current location of the virtual helicopter. Examples of ‘pull mode’ transfer: • The DICE interface polling the infospace to find out the type of a given object, • A Maptek Carmen process starting up and joining the infospace, pulling the total current state of the situation from the infospace in order to synchronise its display with the other workstations. The software used to achieve this connectivity was developed in-house by DSTO. It was written in the programming language Java, using remote method invocation (RMI) to maintain connectivity between the various computers. Connections were established through the use of a “heartbeat channel”, which was implemented using multicast UDP sockets. In addition, an Access database was set up as an archive for the infospace – storing all messages that were transmitted. This facilitated synchronisation of new processes as they came online by allowing them to pull data from the database (as the Carmen system did). Pushing of mission plans was also possible using the Maptek Carmen system. However, the transport mechanism used for the flight plans was not the same as that used for pushing of detections and threat zones, and therefore didn’t have a lot of the benefits associated with the infospace architecture (application independence, robustness, etc). Planned future developments of the infospace architecture are: • Switching from the Java-specific RMI interface to the language-independent CORBA, • Conversion of the Access database log to a more robust multi-user system such as Oracle, • Usage of Oracle’s built-in replication functions to achieve the required robustness, • Evolving smarter pushing and pulling algorithms rather than the “send everything to everyone” approach currently in use, • Possible conversion of the flight plan push information to use the infospace method of communication rather than the specialised method currently in use in order to take advantage of the features of the infospace approach.
insertion of military symbols to denote own force and enemy elements; • depiction of threats with grided, hemispherical threat domes; • textual annotations; • CAD functionality to draw mission profiles; • interfaces to the ‘infospace’ (see below). The Carmen architecture ensures that changes to the position and orientation of entities such as the helicopters are automatically updated in the display modes whenever these are inserted in the Carmen data engine via the ‘infospace’. The effect is of moving models of the helicopters over the 3D terrain display as the helicopters are piloted through the virtual world of the ARH Sim. 4. 3 The Exa-min System The Exa-min sytem is based on the company’s Images+ product which interfaces to the third party MapInfo GIS product. Customised features of the system for Phoenix were the links from the 2D spatial displays to intelligence and surveillance information. One main instance of this was the spatial tagging of video imagery from a UAV demonstrator and the replay of the imagery by ‘point and click’ on the map displays. The Exa-min system was also connected to the ‘infospace’. 4. 4 The Virtual Information Space The situation awareness systems were connected together to form a Virtual Information Space (infospace). The infospace is a conceptual pool of information. The information pool is global and hence accessible to all software that is connected to the infospace. Workstations wishing to connect to the infospace need no prior knowledge of the structure of the network. The infospace was designed for robustness. This was achieved using an “amorphous” structure. In the event that a group of computers are severed from the rest of the infospace (through a network fault, for example), they will continue to function as an infospace unto themselves. When the network connection is restored, it is possible (with some user intervention) to consolidate the information between the two sub-spaces to form a single, consistent infospace. It is anticipated that such a reconciliation process be automated in a future revision of the software. Data transfer through the infospace occurred via one of two modes – the ‘push mode’ (transfer initiated by the supplier of the data) and the ‘pull mode’ (initiated by the consumer of the data).
5.
Scenarios
The general scenario of Exercise Phoenix was the defence of Tindal Air Base against attacks by enemy
Special Forces. Several particular scenarios were developed to involve response of an ARH. Two of these were inserted into the Master Events List (MEL) of the Exercise and four others were played out during the exercise as separate experiments. Because of the restricted ‘play-box’ imposed by the available terrain data set for the virtual environment, it was difficult to devise scenarios that appeared to be part of the enemy activity occurring in the real exercise. As a result, the decision to task an ARH to respond to our particular scenarios was artificial, in that the decision maker generally knew that it was an ARH specific event. (On some occasions though, staff in the Task Force Headquarters did check with DSTO staff present that the events were ARH specific.) Despite this extra artificiality, once the decision was taken to respond to an event with the ARH, the rest of the tasking and mission conduct procedures were adequately exercised. Particular scenarios started with some detections of enemy activity. These detections were variously: • observations of suspicious activity reported by the civil community, giving vagueness in positive identification, intent and location (due to time lapse in reporting and deciding to respond); • detections by Multi Mode Radar (MMR) routine surveillance (no identification and some vagueness in position due to time lapses); • detections by ‘blue’ patrols and observation posts, which gave positive identification and reasonable location accuracy (due to minimal time lapses). Some manual fusion of information by people in TFHQ is carried out to prepare the ‘red’ picture with enemy locations and threat domes marked on the map displays. Some assessment of enemy course of action would also be done and a mission task prepared for the ARHs. Collaborative planning to satisfy this task could then occur involving both TFHQ and MRAB staff. In one example, the plan was first devised at TFHQ and then made available over the infospace to the MRAB, where appropriate staff would review and perhaps modify the plan. Modifications would be made available (also over the infospace) to the TFHQ for their concurrence. All this collaborative planning would occur in real time with both HQs working with the same information. The infospace communication could also be supplemented by voice communications if available. When planning was complete, the SA displays would be used to brief pilots at the MRAB-HQ. As the missions were flown, the pilots would make use of the SA displays: • to navigate to the target area: • to have awareness of the positions of themselves and the second helicopter; • to remind themselves of known blue and red dispositions as briefed previously; • to be advised of new threats. 6.
Outcomes
There was unanimous agreement among the Army personnel involved that the SA concepts demonstrated
would be of considerable value in a real operation. In fact, the Exa-min system has been retained by 1st Aviation Regiment and is being used, with some additional enhancements, as an interim operational capability. The value placed on the system in the helicopter cockpits, by the crew, was higher than expected, as prior opinion was that the helicopter crew would not have enough time to be able to utilise the SA functionality. Indeed the pilots often commented on the desirability of a second display in each cockpit for the pilot’s use. The several virtual missions conducted showed several aspects of the operation of ARHs that require further investigation and these will be included in the proposed demonstration for Ex Crocodile 99. 7.
Enhancements The extensions to the concept demonstration SA system are to address the general areas of user interface and to assist with the target acquisition problem. It is intended that the system be useable by Army staff without the need for extensive training. To this end considerable effort will be devoted to development of an intuitive GUI. One particular area is the preparation of detailed terrain data with draped imagery for virtual ‘fly throughs’ for mission rehearsal. With a large area of operations such as was used at Phoenix it is not possible to pre-prepare terrain data with the required resolution for the whole area, and the eventual ‘playbox’ selected was overly restrictive with respect to the actual Exercise. The intent for the extended system is to prepare the detailed data only as needed. For example, once a target area has been selected (say 3x3kms) from a low-resolution map display, the detailed terrain would be obtained (from some storage medium), the triangulation prepared and then the imagery retrieved and (manually) registered to the triangulated terrain. This operation would be mostly automatic with manual intervention for the registration and should be achievable in just several minutes. Some decision support aids will also be incorporated, in particular the prediction of enemy COA from the prepared enemy picture. This COA prediction is the subject of two currently contracted research projects (with AAII and CSSIP), and is the focus for the data/information fusion thrust of the SA system. Reliable COA prediction will greatly reduce the search area for target acquisition. Additionally, sensor modelling will be incorporated, and in particular, line of sight calculations to show terrain masking effects, to enhance mission planning and to show during missions which areas have been searched and which have been masked. Knowing the availability of assets and deciding which to deploy to satisfy information requirements is another decision support aid which is the subject of additional contracted research (with AOS). It is also planned to implement the Information Management schemes described in another paper at this conference (Unewisse, M., Gaertner, P., Grisogono, AM., and Seymour, R.S., Land Situation Awareness for 2010) as an extension of the infospace which was used at Phoenix. Some additional operational concepts, which this enhanced SA system will enable to be explored, include the possibility of
teaming of assets, such as ARHs and UAVs, and coordination with ground based assets. Shared situational awareness will enable coordination of all elements in an integrated, cohesive, surveillance and response system for a more effective Land Force. 8.
Acknowledgments
Many people contributed to the success of the demonstration at Phoenix but special thanks go to Roger Flint of Maptek, Ken Moule of Exa-min and Captain Wayne Gerrard of the 1 Aviation Regiment.
References 1.
Bowden, F. D. J., Gabrisch, C., and Davies, M., Simulation of Air Defence C3I using the Distributed Effectiveness (DICE) Simulation, Proc. SimTecT97, pp71-76, (1997).
Author Biographies Robert S. Seymour Robert Seymour joined DSTO in 1972 after completing a PhD in Solid State Physics at the University of New South Wales. His early career was in the area of electro-optic materials and lasers and other optical devices with major contributions to the DSTO crystal growth and characterisation facility, theory of solid state lasers and the LADS filter. More recently he conceived and has been involved in the development of the Imaging Laser Radar and has worked in modelling of surveillance systems. He is currently Head of Systems Concepts Discipline in LOD and task manager of the Land Situation Awareness Pictures task. Current interests include C4ISR systems, synthetic environments, land mine detection systems and battlefield surveillance systems, particularly imaging laser radar and unattended ground sensors.
Anne-Marie Grisogono Anne-Marie Grisogono gained her PhD in Mathematical Physics at the University of Adelaide and spent several years in Academia at various overseas institutions, Flinders University and most recently in the Optics Group at the University of Adelaide. She joined DSTO as a Senior Research Scientist in 1995 and has worked on Electro-optic system performance modelling, land mine detection and Land Situational Awareness. In early 1998 she was promoted to Principal Research Scientist and currently heads the Synthetic Environment research activities in LOD.
Jeremy Krieg Jeremy Krieg completed a Bachelor of Electrical and Electronic Engineering from the University of Adelaide in 1997, obtaining First Class Honours. During this time he also worked part-time as an Access database developer for the University of South Australia, in addition to pursuing extra studies in Physics. He joined Systems Concepts Discipline in LOD, DSTO as a Professional Officer Class 1 in January of 1998. During his brief time there he has worked on several diverse tasks including the Battlelab, Imaging Laser Radar the Land Situation Awareness (C4ISR) tasks. His prime area of work during this time has been in the technical development and planning for the Land Situation Awareness task.
