Spectrum

ASA-SimaaS: Advancing Digital Transformation through Simulation Services in the Brazilian Air Force

Joao Dantas — 2025-09-23

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.

Analysis of Computer Vision Application in The Context of SAR Missions

João Custódio de Faria Filho — 2025-09-23

This work includes a general feasibility and performance analysis of some image classification models for Search And Rescue (SAR) operations. Since SAR missions are typically high-wear and high-risk situations for both victims and SAR crew, some teams around the world have been seeking to use technology to speed up rescues and reduce damage. The Croatian Mountain Rescue Service (CMRS) and other SAR teams have been using Unmanned Aerial Systems (UAS) to obtain bird's eye captures of the search area, contrasting with typical low-altitude manned flies. This paper uses images from the HERIDAL dataset to train, validate, and test models. We used two different neural network architectures and eight different training parameters. The accuracy above 98% was achieved, but it does not necessarily mean that the models are appropriate for use in real life, so several considerations were made.

Exploration and Rescue of Shipwreck Survivors using Reinforcement Learning-Empowered Drone Swarms

José Fernando Brancalion — 2025-09-23

The goal of this project is to create a reinforcement learning algorithm that locates shipwrecked individuals using a swarm of drones. A simulated environment was developed to train and visualize the outcome of the trained algorithm, considering the ocean's dynamic circumstances. This project does not discuss image recognition of shipwrecked people, since the true focus of this project is to optimize the search routine of a drone to find the target in the most efficient way possible. The implemented Reinforce algorithm takes into account a dynamic map of probabilities, representing the chances of a person being found, as well as the position of other agents. Outcomes include an open-source python package for the environment and the implementation of the reinforcement learning algorithm. The algorithm demonstrates superiority over the predefined approach, proving the advantages of reinforcement learning in efficiency and effectiveness.

Defining Evaluation Criteria for Strategic Project Portfolio Selection in Brazilian Air Force

Gustavo Barbi Vieira — 2025-09-23

This paper presents a qualitative value model approach for Project Portfolio Selection (PPS) applied to Brazilian Air Force (FAB) projects. It begins with a brief literature review on PPS and Value-Focused Thinking (VFT). The study identifies key objectives from FAB's strategic planning to support the development of a qualitative value model for PPS. It then demonstrates the model through its application to a subportfolio focused on Anti-Aircraft Systems within FAB. Finally, the paper discusses methods for executing the PPS process using the VFT model. This approach innovates by tailoring the value model to the specific decision context of each subportfolio, which enhances the assessment of potential outcomes. Furthermore, it outlines procedures for aggregating PPS results across subportfolios, accounting for interdependencies among projects from different subportfolios.

Comparative Performance Analysis of Two Multi-Aerial Threat Evaluation Algorithms

Humberto Baldessarini Pires — 2025-09-23

In modern aerial defense operation, the evaluation of potential threats is of paramount importance for effective response strategies, particularly when such assessment is performed in real-time. This study presents a comparative analysis of an algorithm developed by the authors, and referred to as DM, and a Markov chain-based approach (MC) in terms of prediction accuracy, execution time, and processing capacity. Notably, DM consistently achieved higher accuracy until simulation time 1350, despite both methods utilizing the same Artificial Neural Network architecture. Additionally, DM exhibited superior execution time and processing capacity, handling a maximum of 89 threats within a one-second timeframe, while MC processed 10 threats. Based on this, it can be asserted that DM meets the requirements for real-time threat evaluation. The results can be attributed to DM's simplified methodology, enabling more accurate and distinct predictions.}

Reliability and Predictive Maintenance of PT6 Engines Using Machine Learning (Random Forests) within the Industry 4.0 Framework

Juan Brazalez — 2025-09-23

Aircraft engine reliability is critical, particularly in military operations where mission success and safety depend on optimal engine performance. The PT6 engines, used in Super Tucano aircraft by the Ecuadorian and Brazilian Air Forces, are renowned for their versatility and robustness. However, their operational demands necessitate advanced maintenance strategies to prevent failures, enhance safety, and minimize downtime. One of the challenges in developing such strategies lies in managing the uncertainties inherent in engine performance and degradation. Variations in operating conditions, environmental factors, and measurement noise introduce uncertainties that can complicate the prediction of failures and the estimation of Remaining Useful Life (RUL). This study addresses these challenges by incorporating a parametric analysis within the machine learning framework, specifically using Random Forests. This approach not only captures the complex relationships between operational parameters and engine degradation but also evaluates the sensitivity of predictions to variations in key inputs. By leveraging Industry 4.0 technologies, including Big Data Analytics and IoT, the study aims to enhance the robustness of predictive maintenance (PdM) models, ensuring operational readiness and cost-effectiveness in both military and civilian aviation contexts.

Inconsistency detection methods for statecharts and sequence diagrams: a systematic literature review

Matheus Cogo — 2025-09-23

During model-based systems engineering or software engineering activities, diagrams representing use cases (sequence diagrams) and diagrams representing object behaviors (state machine diagrams or statecharts) can conflict with each other in what is called an inconsistency. Detecting these inconsistencies is crucial to check if a given specification is realizable through the behavior that was conceived to meet it. This paper provides a systematic literature review of inconsistency detection methods for UML state machine diagrams and sequence diagrams. The selection process is aided by an open-source machine-learning tool, and resulted in the qualitative synthesis of 27 works. The included publications offer methods to tackle the detection of horizontal-semantic behavior inconsistencies.

Quantum Remote Sensing Applied to National Defense

Jullyano Lino da Silva — 2025-09-23

This work presents a study on quantum remote sensing applied to national defense using RADAR (Radio Detection and Ranging) and LIDAR (Light Detection and Ranging) technologies, which can be adopted in metrology and remote sensing processes, capable of detecting and intercepting targets in a given coverage area. In this context, the work presents a bibliographic research addressing concepts, principles of quantum mechanics, basic configurations, methods of photon entanglement, projects, experiments, and potential applications of RADAR and LIDAR systems, highlighting advantages and aspects, such as speed, greater precision, lower cost, greater security guarantee, higher sensitivity, and resolution of quantum remote sensing operation, as well as associated challenges: photon loss, technical difficulties, the phenomenon of decoherence, sensitivity to noise, and the need for integration of quantum remote sensing systems with quantum error correction modules.

Detection of Drone with Convolutional Neural Network

Vinícius Ormianin Arantes Sousa — 2025-09-23

The increasing use of drones is notable in both military operations and various civilian activities. However, the difficulty in detecting these devices has become a concern when it comes to protecting sensitive areas from unauthorized drone flights. Compounding these challenges is the ability of drones to fly at night, adding an extra layer of difficulty to surveillance and information security efforts. This paper explores the use of an expanded CO₂ laser beam, in a laboratory setting, as an illuminator directed at a drone flying in a controlled environment, aiming to capture images in the long-wave infrared (LWIR) spectrum. The acquired images were used to train a convolutional neural network (CNN) using the YOLO (You Only Look Once) architecture. The results demonstrate the feasibility of using this approach to detect drones when illuminated by an energy source.

Modeling and Experimental Evaluation of Amplitude Error in the Determination of the Angle of Arrival in RWR Sensors

Adônis Virgílio Teixeira Punti — 2025-09-23

The Angle of Arrival (AOA) estimated information in Radar Warning Receiver (RWR) sensors using the amplitude comparison method is susceptible to various error sources that can affect its accuracy and precision. Variation in Signal-to-Noise Ratio (SNR), which is directly related to the signal amplitude received by the sensor, is one such factor that can compromise AOA determination. This paper presents a laboratory experiment conducted to assess the AOA error due to variations in SNR. A theoretical analysis is performed, where the SNR is a function of the threat detection angle and the receiver response, including the antenna pattern and position. The experimental analysis validates the theoretical results of the model. Finally, the feasibility of evaluating AOA processing of RWR sensors in a reduced chain configuration through conducted tests is demonstrated, offering an alternative to field or anechoic chamber testing.

Detection Capability of the X-47B UAV Aircraft Based on Static RCS Prediction

Newton A. S. Gomes — 2025-09-23

The doctrinal pillar "Combat Survival" is gaining increasing notoriety in the military environment. Recent conflicts, such as the Ukrainian War, as well as the clash between Azerbaijan and Armenia, indicate the increasing use of loitering munitions and UAVs (\textit{Unmanned Aerial Vehicles}) as aerial attack platforms, reducing human contact on the front lines. This work performs the radar prediction of the X-47B UAV using electromagnetic simulations based on its static RCS (\textit{Radar Cross Section}) in the VHF, L, S, C, and X frequency bands. A hypothetical radar evaluates the greatest detection ranges among the aforementioned bands. The results demonstrate that the VHF band obtained the greatest detection range, with a gain of 272.2$\%$ about the L band. Thus, the RCS analysis of an aerial platform with a delta-shaped flying wing geometry, such as the X-47B UAV, without electromagnetic absorbing materials, does not present stealth characteristics in the analyzed electromagnetic spectrum bands.

A Laboratory Radiometric Calibration of an Electro-Optical Sensor

Marcus V. de Q. S. A. Costa — 2025-09-23

Remote sensing imaging satellites play a vital role in Intelligence, Surveillance, and Reconnaissance (ISR) missions, as they enable the acquisition of information from virtually any location on the Earth’s surface. To ensure the reliability of the provided information, it is essential to calibrate the onboard sensors on these satellites. This paper aims to present a methodology for spectral and radiometric calibration of a Parrot Sequoia camera in laboratory settings. This camera features four monochromatic sensors and one RGB sensor, similar to those onboard orbital platforms. The methodology described employs equipment available at the Laboratory of Radiometry and Characterization of Electro-Optical Sensors (LaRaC) at the Institute for Advanced Studies (IEAv). The paper presents the Spectral Response Functions (SRFs) of the camera sensors, as well as the Radiometric Calibration data. It is worth noting that the proposed methodology can be replicated for any other orbital electro-optical imaging sensor.

Electromagnetic Interference Between Naval Employment Radars

Newton A. S. Gomes — 2025-09-23

The use of radars in the maritime environment has both civil and military applications. The effect of co-channel interference between radars deployed on different ships occurs frequently in several navies. This work proposes to develop a methodology to mitigate the effect of interference between radars operating at the same frequency. A hypothetical radar is used to simulate the effect of co-channel interference between ships, varying the azimuthal engagement angle and the distance between the ships. In the end, a topology of arrangement of ships in a column is obtained, with a maximum distance of 5.95 NM and an arrangement in line with a maximum value of 37.01 NM. This work shows that interference occurs between radars, being overly dependent on the angular condition of engagement due to the effect of the RCS (Radar Cross Section) of the interfering ship.