 |
 |
 |
 |
| |
|||||
|
|
|
The Attack Operations Decision Aid (AODA) is a prototype software tool under development by a team of three contractors: TRW, Logicon, and Alphatech. The main function of the tool is to determine and recommend, in near real time, an optimal assignment plan for diverting assets from pre-planned Air Tasking Order (ATO) missions to Time Critical Targets (TCTs). The goal of the development effort is to have the decision aid interface with the Theater Battle Management Core Systems (TBMCS) as a mission application. The tool is also envisioned to be an integral part of Dynamic Battle Management (DBM) and Joint Target Execution (JTE) concepts, for which AODA would be deployed on airborne platforms such as AWACS and JSTARS. MITRE's support in this program as part of the overall Theater Missile Defense (TMD) Battle Management (BM)/C4I project includes technical requirements document (TRD) development, technical evaluation of competing proposals, review of software architecture and design, evaluation of problem formulations, recommendation of solution algorithms, and monitoring of development progress. AODA will enable the battle manager to identify and commit rapidly the right assets to attack ground TCTs through assessment of weapon capabilities within the limited timelines available. It will cross-reference the ATO and situation display data/tracks, identify available assets (such as F-15, F-16, or other weapon platforms) capable of attacking the targets, and display the attack options for operator consideration. The time constraints for TCTs (including TEL or transporter-erector-launcher) dictate that the AODA function provide automatic recommendations of best available weapons to conduct the required missions. The AODA application software will give the battle manager the best recommendation while also having alternative options available. It will have human-in-the-loop features. AODA can run continuously in the background and replan based on the changing battlespace situation. An intrinsic requirement is the ability of the decision aid to incorporate the priorities in TCT prosecution set by Joint Force Commander and Joint Force Air Component Commander into the decision process. To this end, AODA provides a user-system interface to allow the setting of area of operation, divertable ATO mission types, weapon asset types, munitions types, TCT target types, and threat types. It also allows user modification of weapon asset values, target values, and acceptable level for attrition risk. A recently added requirement will further allow the user to exclude targets from AODA processing by geographic location or area, position uncertainty, classification uncertainty, and status. The current AODA development effort is focused on weapon-target pairing along with the necessary input and output interfaces. The core function is to provide near real-time recommended pairing plans for a decision-maker to adopt or modify. While generating such a recommended plan, the software considers the best route from an attack asset’s location to a TCT in the evaluation of the pair as a possible assignment. However, the route evaluation is only meant to be a feasibility check against any imposed airspace restrictions and the threats in presence; the route is not meant to be used as a detailed flight plan for the asset if the assignment is adopted. The attack asset’s commanding unit is still to provide such a flight plan for execution. To support determining a flight plan that coordinates and deconflicts all assigned missions in space and time, the dynamic airspace management tool called 4-Dimensional Airspace Deconfliction (4DD) decision aid could be used. (See previous article.) When updated status on TCTs, threats, and weapon assets is received by AODA, the application software will start a pairing process (Figure 1). The process includes the following steps:
Valuation of Targets and Assets: A target’s value consists of two components: intrinsic value and dynamic value. Intrinsic value reflects the campaign-wide impact of the survival of the target, while dynamic value represents the immediate capacity of the target to inflict damage on friendly assets. The value of an available weapon asset also consists of intrinsic and dynamic components. The intrinsic component includes the campaign-wide value of all strike and support airframes in the mission and the dynamic component captures the opportunity cost of diverting the weapon asset from the originally planned ATO mission to a TCT prosecution. Evaluation of All Possible Weapon-Target Pairs: The value of a weapon-target pair if taken as an assignment is the net of an adjusted target value minus an adjusted weapon asset value. The adjusted target value is the target value (defined earlier) multiplied by the probability of target kill by the weapon asset. The probability of target kill is in turn a function of the probability of weapon asset’s safe ingress to the target area, the probability of target acquisition by the asset’s on-board sensors, and the probability of target destruction with the weapon load on the asset. For the adjusted weapon asset value, the weapon asset’s intrinsic value (defined earlier) is multiplied by the probability of the weapon asset being killed in the assignment while its dynamic value or the opportunity cost (also defined earlier) will be included in full regardless of the weapon asset being killed or not. The probability of a weapon asset being killed is calculated from the probability of its safe ingress and the probability of its safe egress, both of which depend on the threats present on the asset’s flight route. Determination of an Optimal Pairing Plan: With all possible weapon-target pairs evaluated and each assigned with a value, the pairing optimization will be invoked to determine the best assignment of available weapon assets to TCTs so that the overall value is maximized. The algorithm used for this optimization is a variation of the Auction algorithm developed by D. P. Bertsekas for the problem known in operations research as the assignment problem. User Acceptance or Alteration of Plan: The solution from the optimization will be displayed to the user as a recommended plan for weapon-target pairing. And the associated flight route for each selected pair will be charted on a map of the battlespace along with information on flight time, risks, probability of target kill, and the value of the pair. At this point, the user can decide to take any of the following options: (1) accept and commit to the whole plan, (2) accept and commit to part of the plan, or (3) alter part of the plan to reflect user’s preference not considered in the optimization. Since any user alteration may cause conflicts in the remaining plan, re-optimization with the user alteration as inputs may be needed. Any user commitment will also be honored in subsequent pairing plans. AODA is now in its first 18-month development phase, which is to be followed by a second 18-month effort. It has been selected to participate in EFX99 experiment, which is currently scheduled to occur in August-September 1999. The experiment results will be used to evaluate the effectiveness of the initial prototype application and to modify the second-phase development effort. Results of the prototype development will serve as a basis for full-scale development leading to actual deployment in ground command and control facilities such as AOC and/or on airborne platforms such as AWACS and JSTARS. For more information, please contact Chien-Ching Cho using the employee directory. |
Solutions That Make a Difference.® |
|
|
