edge accuracy and completeness, evaluating operational complexity, and recommending optimum
human-machine decision-making interaction. The
interaction between the human and automated decision space is not illustrated in the simplified concept
shown in figure 12, but this interaction would be significant in tactical operations.
The outputs of the conceptual decision space system would include decision alternatives, estimations
of predicted consequences, estimated probabilities of
success and failure, and the confidence levels associated with source information, options, and knowledge in general.
To coordinate tactical decisions across the force, a
system of distributed decision systems is needed. A
future concept that relies heavily on AI technologies
is to integrate identical intelligent agents onto distributed warfare platforms. These agents would share
data and information and each develop decision
alternatives for both individual resource management and force-level battle management options.
The distributed agents would share decision alternatives and synchronize their selections. This system of
decision systems would enable distributed warfare
coordination with the objective of optimizing warfare resources at the force level. This futuristic concept would depend on intelligent analytical methods
as well as intelligent and self-aware data strategies
and data architectures. Ultimately, such a system of
AI systems would enable huge gains in tactical decision superiority.
Using methods of machine learning to process and
analyze large amounts of heterogeneous data and
information, AI technology can make predictions
about probable effects, outcomes, and responses. These
AI methods, referred to as predictive analytics (PA), can
provide a powerful capability for tactical decision-mak-
ing. Armed with the knowledge of possible effects and
adversary responses to courses of action, warfighters
can leap ahead in terms of applying longer-term strat-
egy to near-term warfare decisions.
A PA capability enables strategic operations within
the tactical domain — enabling projections of possible consequences and effects of decision alternatives.
Conceptually, PA can develop what-if and if-then predictive scenarios to shape the synthesis of future intelligent decisions and coordinated resource management. PA would identify projected short-term and
long-term effects of different course of action options.
It would enable BMAs to assess these projections and
weigh them as courses of action are selected.
Figure 13 contains some of the notional capabili-
ties of a future PA capability. Given tactical knowl-
edge of the operational situation and warfare assets,
as well as COA options developed by the resource
management capability, PA could assess the conse-
quences of COAs and develop projected future states
of the environment and warfare assets. These projec-
tions would be used to support the selection of COAs
with the most desired consequences. A PA capability
would support tactical actions that best align short-
and longer-term objectives. It could also assess the
possible effects of weather predictions and the avail-
ability, depletion, and projected capability of warfare
resources. It could also assess and predict the overall
readiness, resilience, and warfare capability of a tac-
tical battle group.
A PA capability could employ game theory methods to perform war-gaming assessments to predict
enemy responses to tactical actions. A model of the
adversary’s predicted knowledge, capabilities,
intents, and strategies would have to be developed
and maintained based on our knowledge and predictions of the enemy. In addition, it would be necessary to develop and maintain a predicted model of
what the enemy knows about our forces, based on
assumptions and any tactical knowledge we have.
This war-gaming capability could conceptually be
part of the operational BMA capability for tactical
In summary, the battle management problem space
is complex and it will only continue to grow in com-
plexity with the addition of more sensors, more
information, more unmanned threats, more non-
state adversaries, and advances in technology. This
ever-increasing complexity places a greater demand
on tactical decisions — requiring them to be made
both more quickly and more effectively. The level of
complexity can easily exceed the abilities of human
decision-making. Fortunately, the increase in sensors
and information systems is also creating an opportu-
nity for AI as a capability enhancer and improver for
tactical decision support. This paper introduced some
concepts for using AI to improve combat identifica-
tion, shared situational awareness, battle manage-
ment, resource management, and operational war-
gaming. Employing AI effectively will require a
holistic systems of systems approach to create an
adaptive architecture of decision aids that synchro-
nizes distributed knowledge and decisions across the
force, establishes and maintains decision scope, iden-
tifies levels of situational complexity, and self-assess-
es to manage human-machine interaction modes
and determine levels of confidence in knowledge and
COA. The effective use of AI in support of human
warfighters provides the foundation for tactical solu-
tions and decision superiority.
1. A response US Secretary of Defense Donald Rumsfeld provided at a US Department of Defense news briefing, February 12, 2002.