Joao P. A. Dantas

This work explores the use of military simulations to predict and evaluate the outcomes of potential scenarios. It highlights the evolution of these simulations and the increased capabilities driven by advancements in artificial intelligence. Also, it discusses the various applications of military simulations, such as developing tactics and employment doctrines, training decision-makers, evaluating new acquisitions, and developing new technologies. The paper focuses on the Brazilian Air Force's efforts to create its own simulation tool, the Aerospace Simulation Environment (Ambiente de Simulação Aeroespacial — ASA in Portuguese), and how this cloud-based service, called ASA Simulation as a Service (ASA-SimaaS), can provide greater autonomy and cost-efficiency for the military force. The main contribution of this work is to present the ASA-SimaaS solution as a means of empowering digital transformation in defense scenarios, fostering a partnership network, and improving the military's simulation capabilities and competitiveness.

This work presents AsaFG, a human-in-the-loop integration module that connects the Aerospace Simulation Environment (ASA), a constructive simulation environment for military scenarios, with the FlightGear flight simulator. The module enables real-time interaction between human pilots and simulated entities in air combat scenarios, allowing pilots to assume roles within simulated missions and interact with autonomous agents. This integration enables direct interaction between constructive and virtual components, contributing to more realistic and immersive air combat simulations. The work also discusses key challenges during the integration process, including data exchange, control interfaces, and missile modeling differences between the platforms. A specific contribution involves adjusting the missile model in FlightGear to closely match ASA’s implementation, enabling consistent simulation behavior across systems. Validation includes performance metrics and experiments with Brazilian Air Force pilots across three operational scenarios. The results demonstrate the module’s applicability for training, operational analysis, and studies on human-autonomy teaming in simulated air combat scenarios.

This work explores Beyond Visual Range (BVR) air combat simulations, focusing on two-versus-two scenarios involving autonomous agents. The engagement phase in BVR combat presents complex and unpredictable situations, as it is difficult to anticipate the behavior of opposing aircraft and the outcomes of tactical decisions, especially in multi-agent settings. A promising approach is the use of Deep Reinforcement Learning (DRL), which enables agents to learn from dynamic environments. According to fighter pilots, collective situational awareness, defined as understanding the spatial distribution and orientation of allies and opponents, is essential for executing coordinated tactical maneuvers. The main contribution of this work is AsaGym, a library for developing and training DRL-based fighter agents in BVR scenarios. A case study demonstrates its use, applying a reward function that promotes coordination based on collective situational awareness, and compares different DRL algorithms to assess their ability to foster cooperative behavior. The results highlight DRL’s potential to address the complexities of modern air combat and support the development of more adaptive and effective tactics in multi-agent BVR scenarios.

This study presents a methodology for developing models that replicate the complex pop-up attack maneuver in air combat operations, using flight data from a Brazilian Air Force pilot in a 6-degree-of- reedom flight simulator. By applying imitation learning techniques and comparing three algorithms – Multi-Layer Perceptron (MLP), Long Short-Term Memory (LSTM), and Gated Recurrent Unit (GRU) – the research trains models to predict aircraft control inputs through sequences of state-action pairs. The performances of these models were evaluated in terms of Root Mean Squared Error (RMSE), coefficient of determination (R²), training time, and inference time. To further enhance the training dataset with the aim of improving the robustness of the models, a Variational Autoencoder (VAE) was employed to generate synthetic data. These findings demonstrate the potential for deploying such models in fully autonomous aircraft, enhancing autonomous combat systems’ reliability and operational effectiveness in real-world scenarios.

Beyond Visual Range (BVR) air combat is an essential part of modern aerial warfare, relying on advanced radar, missile systems, and decision-support technologies. This survey provides a comprehensive overview of how simulation and Machine Learning (ML) tools have been used to analyze BVR combat, covering key methodologies, applications, and challenges. We examine how ML enables adaptive tactics to improve behavior recognition and threat assessment to enhance situational awareness. The paper also traces the historical evolution of BVR combat, outlining key engagement phases such as detection, missile launch, and post-engagement assessment. A key focus is on the role of simulation environments in modeling realistic combat scenarios, supporting pilot training, and validating AI-driven decision-making strategies. We analyze state-of-the-art simulation tools, comparing their capabilities and limitations for studying multi-agent coordination and real-time adaptability. This survey’s main contributions include descriptions of ML applications in BVR air combat, evaluations of simulation tools, identifications of research gaps, and insights into future research directions. This work provides an overview of how traditional simulation approaches merge with artificial intelligence to enable advanced, effective human and autonomous decision-making systems in dynamic and contested environments.

Aircraft weight is a key input in flight trajectory prediction and environmental impact assessment tools. However, the lack of openly available data regarding the actual aircraft weight throughout the flight requires the development of mass estimation approaches to be incorporated into these tools. This study uses large-scale open aviation data made available by Eurocontrol's Performance Review Commission to develop an open-source machine learning model to predict commercial flights' actual takeoff weight. The data combines detailed flight, trajectory, and meteorological information for 369,013 flights that transited through the European airspace in 2022. Several operational features are created to represent each flight's horizontal and vertical profiles accurately. For model learning, we employ CatBoost, LightGBM, XGBoost, artificial neural networks, and an ensemble of these models, which were selected for their robust performance in structured data analysis and potential for high predictive accuracy. The models are evaluated based on their efficiency, accuracy, and applicability to real-world data. The best-performing model is found to predict the aircraft takeoff weights with a mean percentage of error of 1.73%.

We introduce TartanAviation, an open-source multi-modal dataset focused on terminal-area airspace operations. TartanAviation provides a holistic view of the airport environment by concurrently collecting image, speech, and ADS-B trajectory data using setups installed inside airport boundaries. The datasets were collected at both towered and non-towered airfields across multiple months to capture diversity in aircraft operations, seasons, aircraft types, and weather conditions. In total, TartanAviation provides 3.1M images, 3374 hours of Air Traffic Control speech data, and 661 days of ADS-B trajectory data. In addition to the raw data, we provide post-processed versions with synchronized, filtered, and interpolated data. In addition to the dataset, we also open-source the code-base used to collect and pre-process the dataset, further enhancing accessibility and usability. We believe this dataset has many potential use cases and would be particularly vital in allowing AI and machine learning technologies to be integrated into air traffic control systems and advance the adoption of autonomous aircraft in the airspace.

This work introduces PoKER, a novel probabilistic model engineered to optimize missile launch effectiveness in air-to-air scenarios, specifically within Beyond Visual Range (BVR) air combat. Unlike conventional Weapon Engagement Zone (WEZ) models that delineate zones based on static distances such as maximum, minimum, and no-escape ranges, PoKER applies machine learning to predict kill probabilities more accurately by integrating the stochastic behaviors of targets and missile miss distances. This model dynamically evaluates target behavior, greatly expanding the predictive capabilities of engagement analysis. By factoring in elements such as warhead lethality, target and shooter orientations, and the specific conditions of engagement, PoKER provides important insights into engagement dynamics and quantifies success probabilities. Consequently, it can potentially be an important tool for BVR air combat pilots, improving operational decision-making within this specialized combat domain.