Opportunities and Constraints in Applying Artificial Intelligence in Military Enterprise

Introduction

Artificial Intelligence (AI) stands out as a new magical way to transform the digital age even further heights. Are the Armed Forces, as part of modern society, ready to apply the AI and use it to gain advantages against adversaries or are they unable to benefit from discrete innovations? How can we assess the readiness of military enterprise in adopting or innovating new capabilities enhanced by the AI related technologies?

The short paper uses an enterprise architecture (EA) tool developed specially for military enterprises to assess the opportunities and challenges in adapting the benefits of AI. The EA tool analyses the strategic posture and operational processes of a military force. Furthermore, it focuses primarily on the command and control related capabilities including sensemaking, decision making, and organisational learning. Additionally, the tool helps to analyse the readiness of information, security and technical structures of armed forces.

Theory and literature review

A military enterprise can be defined as open, complex, a socio-technical system that exists in the national and geopolitical environment. The enterprise is evolving gradually being affected by its history,culture, surrounding society, and what opportunities are available for the future. The knowledge-driven evolutionary model is used to compose an EA tool that helps military architects to analyse opportunities and constraints in evolving the military with AI based capabilities.

Using the Thorpe et al. view of the evolution of business knowledge, Mattila and Parkinson define the evolutionary roadmap for strategic posture in confrontation, doctrinal improvement, command and control, and military information management. The supporting technical layers of information security and ICT infrastructure are studied correspondingly and combined in the framework. 

The EA tool merges the above layers and defines the forces acting within the structure presented in Figure 1.

Figure 1: Military enterprise structure from enterprise architecture viewpoint.

A literature survey was made to create an understanding of the current opportunities and challenges that enterprises feel they are facing when considering improving their business with AI enhanced features. The survey was done through an Internet search explained in Table 1. 
Table 1: Parameters of the literature survey on AI implementation

The four areas of AI opportunities and challenges and their eight key issues create the performance metrics for the EA tool in the following section.

Research

The approach of this research is pragmatic since the work intends to anticipate opportunities and constraints when military applies AI to accelerate their C4I transformations. The postulated EA tool is composed of previous work and separate studies using qualitative deduction. The literature review provides a statistical data concerning opportunities and challenges in applying AI in any enterprise. The collected AI data is projected to a case study of an anonymous Armed Forces (Blue Force) C4I structure and its intended improvement. The feasibility of the EA tool is measured by its ability to anticipate the accelerators and obstacles in the journey of AI implementation. Further study is needed to measure the value of the information concerning accelerators and barrier when AI features are being implemented.

The case study of the Blue Force uses the EA tool to make better sense of the whole situation of the enterprise in aiming to apply AI and gain an advantage over the adversaries is illustrated in Figure 2. The Blue Force seems to be gaining advantage against their adversaries from the evolutionary posture since it has been acquiring the modern armament steadily. There may, however, be a tendency towards the operational posture as the cost of contemporary armament is rising and the available workforce is diminishing. 

The Blue Force strategy for military process performance seems to be on a path towards coordination aiming for joint force capabilities. Subsequently, there are also indications towards unified logistics and replicated force generation. 

The essential parts of the Blue Force command and control capabilities are based on learning by drilling and somewhat by understanding.  Furthermore, the Blue Force decision making has an authoritarian approach with a touch of shared intent. The sensemaking seems to focus to the areas of known and knowable. 

The information management of the Blue Force is somewhere between folder and page management, and the information security is mainly based on controls within each domain. There seem to be advanced bandwidth and mobility services available, whereas computing seems decentralised and connected no further than forest level. ICT operations seem to be at system management level.

Figure 2: Military enterprise structure from enterprise architecture viewpoint.

Results and discussion

The specific concerns in AI implementation into generic enterprises are reflected against the results of analysis of the Blue Force enterprise structure. Table 2 is illustrating this reflection either positive, i.e., opportunity or negative, i.e., challenge. The direct guidance is either utilising the opportunities or trying to mitigate the challenges in coming AI implementation within the Blue Force.

Table 2: Testing the EA tool in analysing opportunities and challenges in applying AI within the Blue Force

The EA tool supports the analysis of enterprise from cultural dimension down to technical dimensions covering the indicated areas of concern in AI implementation. Therefore, the EA tool is sufficiently holistic in modelling the military enterprise structure.

The reflection of architectural analysis of the Blue Forces enterprise structure against eight key issues in AI implementation provides the enterprise architect with ten opportunities to accelerate the adaptation of AI and recognises four challenges requiring mitigation. Consequently, the EA tool recognises both driving and hindering forces within an enterprise.

The EA tool identified four inter-layer dependencies (ID) in implementing AI. Consequently, the EA tool recognises inter-dependencies through the enterprise structure compared to layer-oriented models.

Conclusions

The pragmatic research uses evolutionary enterprise architecture tool in analysing the opportunities and challenges in applying artificial intelligence features in a military enterprise.
The proposed EA tool covers the whole area of concern in an AI implementation. The tool recognises both opportunities and challenges that can be addressed in an AI implementation plan. The tool models also more complex inter-dependencies within the enterprise structure.
The EA tool appears to help military enterprise architects in analysing the status of military force aiming to benefit from artificial intelligence features. Nevertheless, this case study and the initial stage does not prove the feasibility of the EA tool entirely. Therefore, there is a need for further study both within the specific case study and possibly broader cases among Armed Forces to improve the EA tool.