GhostPlay's AI Advancements Can Reshape Military Decision-Making and Future Warfare

Photo credits: 21strategies

GhostPlay, a pioneering joint research endeavor involving German AI manufacturer 21strategies, sensor manufacturer Hensoldt, the University of the Federal Armed Forces in Hamburg, and consulting firm Borchert Consulting & Research, has captured the attention of esteemed publication ES&T. In a recent article, ES&T delves into the noteworthy accomplishments of the project and its potential implications for military decision-making processes. GhostPlay's primary objective is to propel the boundaries of an AI-based digital twin of the battlefield, aiming to revolutionize military operations. The Office of Army Development has also expressed keen interest in the project, specifically tasking the developers with exploring the utilization of loitering ammunition deployed by helicopters at the virtual battlefield.

Simulating AI-enhanced ground-based air defense systems

During the initial phase of GhostPlay, which concluded in late 2022, the focus was on simulating scenarios related to the suppression of enemy air defense (SEAD). The simulations involved an air defense group consisting of cheetah ("Gepard") anti-aircraft tanks fending off a drone swarm of over 100 units. To achieve this, the project required the establishment of an appropriate simulation environment and AI training system, along with the formulation of crucial rules and constraints for the scenario. Remarkably, the AI-powered GhostPlay software developed by 21strategies significantly enhanced the defense capabilities of the Gepard tanks, boosting their performance and resilience. Through the employment of a virtually AI-augmented group of digital twin Gepards, the system was able to defend against up to 60 unmanned flying objects, six times the usual limit of a single tank. This breakthrough significantly improved the overall defense capabilities of the simulated air defense system.

Optimizing Swarm Dynamics through emergence

Currently, in the second phase of GhostPlay, the project's developers are collaborating with the Office of Army Development to simulate the deployment of loitering munitions by helicopters. The primary aim is to leverage AI to optimize a swarm of these munitions, enabling them to effectively counter robust enemy air defenses. This phase presented additional challenges, such as accurately representing terrain conditions, buildings, vegetation, and infrastructure details essential for incorporating low-level flight tactical possibilities. Sophisticated terrain models are employed to train and evaluate the tactical behavior and optimal deployment of the virtual loitering munitions swarm.

 

The simulations conducted within GhostPlay go beyond basic swarm behavior, such as formation flying or evasive maneuvers. The project places emphasis on achieving coordinated attack behavior, allowing swarm members to collaborate in achieving mission objectives. The AI system is designed to divide tasks into sub-goals and determine when and how information should be shared among swarm members through direct communication or by interpreting observed behavior. The simulations follow the principle of "counter play," where both the air defense and attacking swarm behavior are continuously refined during each run. This dynamic approach empowers the AI system to develop new tactics and respond adeptly to evolving situations. Notably, the GhostPlay developers believe that the air defense tactics learned through the project could be potentially applied to real-world scenarios, aiding in the repulsion of attacks against Western air defense systems.

GhostPlay's transformative AI advancements can shape future warfare

As the GhostPlay project progresses, the existing simulation model will be expanded to incorporate additional active agents, including anti-radar missiles and sensors. The objective is to enable the swarm to make spontaneous decisions in operational environments, deviating from the original plan to gain tactical advantages. This expansion aims to create a system that accurately simulates a capable adversary and selectively matches or exceeds the performance profiles of the user's own forces.

 

The example of GhostPlay underscores the revolutionary impact AI will have on future military training, operational planning, and battle management. It highlights the significance of advanced AI capabilities for both offensive and defensive systems, which are crucial for success in complex and dynamic warfare scenarios. The potential of AI to enhance system performance and provide readily deployable solutions signifies a transformative shift in modern warfare approaches.

 

The GhostPlay project exemplifies the importance of collaborative efforts and cutting-edge AI technology in advancing military decision-making processes. The evaluation acknowledges the accomplishments and implications of GhostPlay, showcasing its contributions to defense optimization, enhanced resilience, and future applications within the military domain. Notably, GhostPlay is funded and commissioned by the Center for Digitalization and Technology Research of the German Armed Forces (dtec.bw), with its primary objective being the development and testing of AI-based decision-making processes within a virtual battlefield.

 

Source:
Ole Henckel (2023). KI-Gefechtssimulation für die Bundeswehr: GhostPlay. In: Europäische Sicherheit und Technik, 72. Jg., Juli 2023, S. 39-40. Bonn: Mittler Report Verlag GmbH. https://esut.de/2023/07/fachbeitraege/42917/ki-gefechtssimulation-fuer-die-bundeswehr-ghostplay/ (15. July 2023).