A CPM task hierarchy that realises an objective 

Contexts in source publication

Context 1

... also introduces an object type for the description of end states . The desired end state of a campaign should describe an acceptable military and/or political world state that must exist before the campaign can be considered complete. Objectives are derived from this end state and it thus forms the basis of operations planning. [12] As shown in Figure 7, a CPM plan uses the ‘has as endstate’ relationship to identify a particular CPM objective that defines the desired end state of the plan. Note that this relationship originates from a CPM container object (not shown in Figure 7) used to identify the intent of a commander to achieve an assigned mission (these CPM planning artefacts are detailed later in this section). We believe that the CPM sub-structure described above is sufficient to express the meaning of a TOPFAS end state object, but a more in-depth investigation is required to tease out any subtle semantic differences and necessary CPM extensions. Tasks are used in a CPM plan to express that an agent is required to execute an activity in order to meet an objective by causing a change in world state [7]. Planned tasks are assigned by a superior agent to a subordinate, are given permissible timings (its temporal properties are equivalent to those of objectives), and an available pool of resources. That a task is performed in order to meet an objective is captured using the realises relationship. Generally speaking, an objective is realised by a single task (the CPM semantics of alternative tasks realising a single objective are currently undefined). However, tasks can be decomposed into hierarchies of subtasks, necessary (but not sufficient) for their completion, using the has as subtask relationship. The above definitions are illustrated in Figure 8; the tasks OA-01 and OA-02 are subtasks of a single, shared parent and so are necessary for the completion of this parent task. Once complete, the parent task realises the objective ODC-01. The definition of the 'change in world state' that results from the achievement of a task can be captured in the CPM by associating the task over the has as effect relationship to an effect. Effects have temporal properties (equivalent in nature to those of objectives and tasks), and can also be preconditions for tasks [7]. Figure 9 shows how effect and precondition specifications have been used to enrich the model shown previously in Figure 8 with this information. The CPM does not yet formally define the implications of task preconditioning, including (1) the implications of a task having multiple preconditions, and (2) a specification of the temporal constraints a precondition imposes on a task (and visa versa). Our initial (partial) intuitive specification (which we will assume is relevant to military planning for now) with respect to the above is: the set of effects that are preconditions This statement, in conjunction with the CPM constraint stating that a n effect can only hold after the completion of task that causes it , permits us to infer that (as shown in Figure 9) initiating execution of OA-03 must occur after the completion of (the parent task of) OA-01 and OA-02. Based on our interpretation of informal TOPFAS vocabulary definitions, we believe that three TOPFAS entities closely match the semantics of CPM task and effect described above, namely the TOPFAS definitions of action/task 2 and effect. In examples we have seen, both CPM tasks and TOPFAS actions are typically associated with an imperative statement starting with a verb (e.g. “seize ...”, “destroy...”, “defeat...”), and associated effects are typically declarative statements that describe the (expected) resulting world state (in much the same way as objectives). Moreover, like CPM tasks, TOPFAS actions can be hierarchically decomposed, and we believe that the meaning of the hierarchical link used is equivalent to that of (the inverse of) the has as subtask CPM relationship. As with objectives, there is an evident distinction in the way the models capture the temporal extents of actions and effects; in contrast to the CPM, TOPFAS actions do not permit the specification of multiple permissible intervals of time, and TOPFAS effects are associated with two distinct time intervals indicating that they are either 'building-up' or 'being maintained'. Additionally, we have not yet identified (or devised) CPM analogues for various TOPFAS action/effect attributes, including action metrics such as willingness, risk level , and efficiency , or those that distinguish between implied and assigned actions. Finally, in the CPM, a task is related directly to the objective it realises, and an effect is related to the objective(s) it supports indirectly via a chained effect → preconditions → task → realises → objective relationship. TOPFAS adopts the opposite approach; effects are directly asserted to support objectives (although this is not included in the example TOPFAS plan shown earlier in Figure 5), and actions provide support via effects. We believe that these both effectively capture the same essential meaning, but a formal investigation of this issue will nevertheless be necessary. So far, we have covered objectives, end states , actions and effects . A missing (but fundamental) component of any military plan is a definition of several lines of operation that specify how an unacceptable situation in the world can be sequentially transformed into an acceptable one (i.e. the desired end state). In basic terms, this situational transformation occurs through the execution of actions to create effects. These effects provide the concrete support for a hierarchy of objectives (operational, to military strategic to NATO strategic) which culminates in the realisation of the overall desired end state [12]. An important component of a plan at the operational level is the 'operational design'. One of the purposes of this artefact is to establish decisive conditions along lines of operation. Our current understanding is that a decisive condition, since it describes a desired world state (ultimately) necessary for the attainment of an overall desired end state, is similar in meaning to an objective. The primary distinction, we believe, is that a decisive condition describes a world state identified by a military commander as essential to containing or neutralising an opponent's centres of gravity and protecting one's own. As shown earlier in Figure 4, decisive conditions in a TOPFAS plan are typically related to one another over a line of operation. We believe that, together, these are used in TOPFAS primarily for operational sequencing: to specify the causal (and thus temporal) constraints between decisive conditions, and describe the ordered dependencies ...

Context 2

... also introduces an object type for the description of end states . The desired end state of a campaign should describe an acceptable military and/or political world state that must exist before the campaign can be considered complete. Objectives are derived from this end state and it thus forms the basis of operations planning. [12] As shown in Figure 7, a CPM plan uses the ‘has as endstate’ relationship to identify a particular CPM objective that defines the desired end state of the plan. Note that this relationship originates from a CPM container object (not shown in Figure 7) used to identify the intent of a commander to achieve an assigned mission (these CPM planning artefacts are detailed later in this section). We believe that the CPM sub-structure described above is sufficient to express the meaning of a TOPFAS end state object, but a more in-depth investigation is required to tease out any subtle semantic differences and necessary CPM extensions. Tasks are used in a CPM plan to express that an agent is required to execute an activity in order to meet an objective by causing a change in world state [7]. Planned tasks are assigned by a superior agent to a subordinate, are given permissible timings (its temporal properties are equivalent to those of objectives), and an available pool of resources. That a task is performed in order to meet an objective is captured using the realises relationship. Generally speaking, an objective is realised by a single task (the CPM semantics of alternative tasks realising a single objective are currently undefined). However, tasks can be decomposed into hierarchies of subtasks, necessary (but not sufficient) for their completion, using the has as subtask relationship. The above definitions are illustrated in Figure 8; the tasks OA-01 and OA-02 are subtasks of a single, shared parent and so are necessary for the completion of this parent task. Once complete, the parent task realises the objective ODC-01. The definition of the 'change in world state' that results from the achievement of a task can be captured in the CPM by associating the task over the has as effect relationship to an effect. Effects have temporal properties (equivalent in nature to those of objectives and tasks), and can also be preconditions for tasks [7]. Figure 9 shows how effect and precondition specifications have been used to enrich the model shown previously in Figure 8 with this information. The CPM does not yet formally define the implications of task preconditioning, including (1) the implications of a task having multiple preconditions, and (2) a specification of the temporal constraints a precondition imposes on a task (and visa versa). Our initial (partial) intuitive specification (which we will assume is relevant to military planning for now) with respect to the above is: the set of effects that are preconditions This statement, in conjunction with the CPM constraint stating that a n effect can only hold after the completion of task that causes it , permits us to infer that (as shown in Figure 9) initiating execution of OA-03 must occur after the completion of (the parent task of) OA-01 and OA-02. Based on our interpretation of informal TOPFAS vocabulary definitions, we believe that three TOPFAS entities closely match the semantics of CPM task and effect described above, namely the TOPFAS definitions of action/task 2 and effect. In examples we have seen, both CPM tasks and TOPFAS actions are typically associated with an imperative statement starting with a verb (e.g. “seize ...”, “destroy...”, “defeat...”), and associated effects are typically declarative statements that describe the (expected) resulting world state (in much the same way as objectives). Moreover, like CPM tasks, TOPFAS actions can be hierarchically decomposed, and we believe that the meaning of the hierarchical link used is equivalent to that of (the inverse of) the has as subtask CPM relationship. As with objectives, there is an evident distinction in the way the models capture the temporal extents of actions and effects; in contrast to the CPM, TOPFAS actions do not permit the specification of multiple permissible intervals of time, and TOPFAS effects are associated with two distinct time intervals indicating that they are either 'building-up' or 'being maintained'. Additionally, we have not yet identified (or devised) CPM analogues for various TOPFAS action/effect attributes, including action metrics such as willingness, risk level , and efficiency , or those that distinguish between implied and assigned actions. Finally, in the CPM, a task is related directly to the objective it realises, and an effect is related to the objective(s) it supports indirectly via a chained effect → preconditions → task → realises → objective relationship. TOPFAS adopts the opposite approach; effects are directly asserted to support objectives (although this is not included in the example TOPFAS plan shown earlier in Figure 5), and actions provide support via effects. We believe that these both effectively capture the same essential meaning, but a formal investigation of this issue will nevertheless be necessary. So far, we have covered objectives, end states , actions and effects . A missing (but fundamental) component of any military plan is a definition of several lines of operation that specify how an unacceptable situation in the world can be sequentially transformed into an acceptable one (i.e. the desired end state). In basic terms, this situational transformation occurs through the execution of actions to create effects. These effects provide the concrete support for a hierarchy of objectives (operational, to military strategic to NATO strategic) which culminates in the realisation of the overall desired end state [12]. An important component of a plan at the operational level is the 'operational design'. One of the purposes of this artefact is to establish decisive conditions along lines of operation. Our current understanding is that a decisive condition, since it describes a desired world state (ultimately) necessary for the attainment of an overall desired end state, is similar in meaning to an objective. The primary distinction, we believe, is that a decisive condition describes a world state identified by a military commander as essential to containing or neutralising an opponent's centres of gravity and protecting one's own. As shown earlier in Figure 4, decisive conditions in a TOPFAS plan are typically related to one another over a line of operation. We believe that, together, these are used in TOPFAS primarily for operational sequencing: to specify the causal (and thus temporal) constraints between decisive conditions, and describe the ordered dependencies between actions, effects, decisive conditions and objectives. To capture decisive conditions and lines of operation in the CPM we make use of the mapping exemplified in Figure 9 above. Decisive conditions are mapped directly to CPM objectives (as with ODC-01 and ODC-02). To capture the sequencing (we believe) is implied by a line of operation, we make use of the aforementioned effect → precondition → task relationship. In Figure 9, the tasks OA-01 and OA-02 realise (via a parent task) the 'decisive condition' ODC-01. The effects of these tasks (OE-01 and OE-02) are both asserted to be preconditions of the task OA-03, which realises ...