The GRACE Consortium

MCFP

Manoeuvre coordinator and formation planner

Military tools provided
Novel techniques used
Example of use
Software overview
Evaluation
Exploitation

Military tools provided

The main objective of the MCFP facility is to support the construction of a route plan for a Task Group. This implies a route plan for all Task Elements of this Task Group. The benefits to the OTC lie in the quality of the route plans, inclusive the consistency between plans. The quality of the route plans is defined in terms of accuracy in timing and known minimum or maximum values of route or platform parameters.

The MCFP facility provides services for the planning and synchronisation of routes. The OTC can verify the feasibility of certain manoeuvres and monitor them during the mission execution. The OTC defines the problem in terms of routes and constraints (required or maximum speed, arrival/departure/linking times) and MCFP automatically determines a solution, if it exists, or the absence of a solution. In case a solution does not exist, the OTC can relax the constraints and look for a partial solution.

This facility can be used in the mission preparation and planning phase, as well as during mission execution. It has been demonstrated when the landing force approaches the coast to determine the best route to enter the fjord and the best splitting point for the SAG.

Novel techniques used

The MCFP facility provides automated services to support constructing and critiqueing route plans, services which are hardly available in current command and control systems. It provides services which are difficult or costly to implement using more "traditional" processing techniques, and makes them available such that they can be used fruitfully by other facilities.

MCFP deals with solving problems of coordination in space and time as a part of route planning, problems which are in general much harder to solve by other means than a constraint solver that is by its nature dedicated to solving this kind of problems.

The Manoeuvre Coordinator and Formation Planner (MCFP) facility was designed and implemented as a multi-agent application, MCFP Calculator and MCFP Demonstrator, that interacts with the DOHP as well. This work successfully demonstrates the combination of multi-agent technology and constraint solving techniques applied to a specific military application.

Example of use

The following example is a small part of the examples used to demonstrate MCFP, and can be found in PD49.2a as well as in W412.B.

The OTC decides to give SAG1 coverage by HAG1 when in the vicinity of the isle of Karmoy at the Northern edge of the Boknafjorden entrance. That coverage must effective along route R05 from waypoint w53 to w54, and along route R02 from w23 to w24. During the covering action, the SAG1 speed shall be 18 knots.

Use is made of MCFP by defining the start time of the operation and deriving its end time with the help of MCFP. The start time is 8.00 GMT May 18, while the end time must be between 8.45 GMT May 18 and 9.15 GMT May 18.

Operations table for Extension 2

Coverage of SAG1 by HAG1

Software overview

Agent design

MCFP has been implemented as two agents. The first agent MDEM serves the demonstration of the second agent MCALC. The agent MCALC offers the services Calc1R, Calc2R, and CalcLR. The MDEM agent makes use of these services of MCALC. For details see next.

Service design

Calc1R calculates the feasibility of a single route that is subjected to constraints on speed and on time.

Calc2R calculates the feasibility of two routes that are linked by constraints, while each route is subject to constraints on speed and on time as well.

CalcLR calculates the feasibility of two routes that are linked by a common point, called fork point, and that are also subject to constraints on speed and on time.

All three services include their own HCI component for interaction with the user of the facility.

Evaluation

The main remark to be made on this facility is that in typical planning situations this kind of problem is not so difficult as to justify the use of a powerful software tool to support its solution. The requirements for synchronisation are not so strong as to require the search of an optimum solution and the OTC can simply adapt the current plan if a significant divergence from the expected result (e.g. reducing speed if he forecasts an anticipated early arrival). Taken together, the plan checking/monitoring use of the facility appears to be more likely than its use during the planning phase.

It is however believed that there is anyhow a need for tools supporting kinematics computations and time/space reasoning. Moreover, the availability of advanced tools such as the MCFP facility can accelerate the identification of requirements for a stronger support in situations when it is necessary to have a very precise temporisation and synchronisation or where very strong constraints in speed (e.g. adverse weather conditions, or low fuel consumption) exist.

To summarize, this facility provides a new tool to help define the best route when there are one or more units and different possible solutions to achieve the target. Using this facility the OTC can find the best routes satisfying unit constraints like maximum speed, or fuel consumption, and then obtain the maximum result from his plans. Even though present operational requirements for route planning seem to suggest simpler functionality, it is foreseen that the more complex functionality provided by MCFP can be useful in critical situations (time pressure, need for precise synchronization etc.). For example, the OTC, before starting the mission, defines different strategies to approach the coast and land, using MCFP; during the mission it could be necessary to switch from one strategy to another minimizing the time, the fuel consumption, etc. In this case the MCFP facility can calculate the best solution (routes) for those different strategies.

Exploitation

The MCFP facility provides services that support a user like the OTC, in exploring and constructing plans for manoeuvring, while various constraints are applied. The delivery contains two agents. One agent, called MCALC, that applies constraint solving techniques and is the actual AI part of MCFP, and another agent, called MDEM, that is used to demonstrate the use of the constraint solving agent MCALC. Special attention has been given to the HCI aspects of the used constraint solving technique. Also note that the agent MDEM contains the domain knowledge, while the constraint solving agent MCALC is a solver application to support aspects of route planning, and is not dependent on the application domain.

Produced by: Peter Martin / Russell Gordon / Mike Pockney Updated: 30 March 1999 Copyright Logica 1999