arXiv:2512.03729v1 Announce Type: cross 
Abstract: The US Naval Research Laboratory's (NRL's) Autonomous Planning In-space Assembly Reinforcement-learning free-flYer (APIARY) experiment pioneers the use of reinforcement learning (RL) for control of free-flying robots in the zero-gravity (zero-G) environment of space. On Tuesday, May 27th 2025 the APIARY team conducted the first ever, to our knowledge, RL control of a free-flyer in space using the NASA Astrobee robot on-board the International Space Station (ISS). A robust 6-degrees of freedom (DOF) control policy was trained using an actor-critic Proximal Policy Optimization (PPO) network within the NVIDIA Isaac Lab simulation environment, randomizing over goal poses and mass distributions to enhance robustness. This paper details the simulation testing, ground testing, and flight validation of this experiment. This on-orbit demonstration validates the transformative potential of RL for improving robotic autonomy, enabling rapid development and deployment (in minutes to hours) of tailored behaviors for space exploration, logistics, and real-time mission needs.

نجح مختبر الأبحاث البحرية الأمريكية في تجربة APIARY في إظهار أول تحكم باستخدام التعلم المعزز لروبوت طائر حر في الفضاء، باستخدام روبوت Astrobee التابع لناسا على متن محطة الفضاء الدولية في 27 مايو 2025. تضمنت هذه التجربة تدريب سياسة تحكم قوية في بيئة محاكاة، مما عزز قدرة الروبوت على العمل بشكل مستقل في ظروف انعدام الجاذبية.

El experimento APIARY del Laboratorio de Investigación Naval de EE. UU. demostró con éxito el primer control de un robot volador libre en el espacio mediante aprendizaje por refuerzo, utilizando el robot Astrobee de la NASA a bordo de la Estación Espacial Internacional el 27 de mayo de 2025. Este experimento implicó el entrenamiento de una política de control robusta en un entorno simulado, mejorando la capacidad del robot para operar de manera autónoma en condiciones de microgravedad.

L'expérience APIARY du Laboratoire de recherche naval des États-Unis a réussi à démontrer le premier contrôle par apprentissage par renforcement d'un robot volant libre dans l'espace, en utilisant le robot Astrobee de la NASA à bord de la Station spatiale internationale le 27 mai 2025. Cette expérience a impliqué l'entraînement d'une politique de contrôle robuste dans un environnement simulé, améliorant la capacité du robot à fonctionner de manière autonome en conditions de microgravité.

The US Naval Research Laboratory's APIARY experiment successfully demonstrated the first reinforcement learning control of a free-flying robot in space using the NASA Astrobee on the International Space Station on May 27, 2025. This experiment involved training a robust control policy in a simulated environment, enhancing the robot's ability to operate autonomously in zero-gravity conditions.

Autonomous Planning In-space Assembly Reinforcement-learning free-flYer (APIARY) International Space Station Astrobee Testing

arXiv:2512.03736v1 Announce Type: cross 
Abstract: Reinforcement learning (RL) offers transformative potential for robotic control in space. We present the first on-orbit demonstration of RL-based autonomous control of a free-flying robot, the NASA Astrobee, aboard the International Space Station (ISS). Using NVIDIA's Omniverse physics simulator and curriculum learning, we trained a deep neural network to replace Astrobee's standard attitude and translation control, enabling it to navigate in microgravity. Our results validate a novel training pipeline that bridges the simulation-to-reality (Sim2Real) gap, utilizing a GPU-accelerated, scientific-grade simulation environment for efficient Monte Carlo RL training. This successful deployment demonstrates the feasibility of training RL policies terrestrially and transferring them to space-based applications. This paves the way for future work in In-Space Servicing, Assembly, and Manufacturing (ISAM), enabling rapid on-orbit adaptation to dynamic mission requirements.

نجح NASA Astrobee في عرض أول تطبيق في المدار لتعلم التعزيز (RL) للتحكم الذاتي على متن محطة الفضاء الدولية (ISS). وقد شمل هذا الإنجاز تدريب شبكة عصبية عميقة باستخدام محاكي الفيزياء Omniverse من NVIDIA، مما أتاح لـ Astrobee التنقل بفعالية في بيئات الجاذبية الصغرى. تؤكد النتائج على صحة خط أنابيب تدريب جديد يسد الفجوة بين المحاكاة والواقع، مما يظهر إمكانيات RL في الروبوتات الفضائية.

El NASA Astrobee ha demostrado con éxito la primera aplicación en órbita del aprendizaje por refuerzo (RL) para el control autónomo a bordo de la Estación Espacial Internacional (ISS). Este logro implicó el entrenamiento de una red neuronal profunda utilizando el simulador físico Omniverse de NVIDIA, lo que permitió al Astrobee navegar de manera efectiva en entornos de microgravedad. Los resultados validan un nuevo pipeline de entrenamiento que cierra la brecha entre la simulación y la realidad, mostrando el potencial del RL en la robótica espacial.

Le NASA Astrobee a réussi à démontrer la première application en orbite de l'apprentissage par renforcement (RL) pour le contrôle autonome à bord de la Station spatiale internationale (ISS). Cette réalisation a impliqué l'entraînement d'un réseau de neurones profonds à l'aide du simulateur physique Omniverse de NVIDIA, permettant à l'Astrobee de naviguer efficacement dans des environnements en microgravité. Les résultats valident un nouveau pipeline d'entraînement qui comble le fossé entre simulation et réalité, montrant le potentiel du RL dans la robotique spatiale.

The NASA Astrobee has successfully demonstrated the first on-orbit application of reinforcement learning (RL) for autonomous control aboard the International Space Station (ISS). This achievement involved training a deep neural network using NVIDIA's Omniverse physics simulator, allowing the Astrobee to navigate effectively in microgravity environments. The results validate a new training pipeline that bridges the simulation-to-reality gap, showcasing the potential for RL in space robotics.

Crossing the Sim2Real Gap Between Simulation and Ground Testing to Space Deployment of Autonomous Free-flyer Control

arXiv:2512.03257v1 Announce Type: cross 
Abstract: Rapid and accurate wildfire detection is crucial for emergency response and environmental management. In airborne and spaceborne missions, real-time algorithms must distinguish between no fire, active fire, and post-fire conditions, and estimate fire intensity. Multispectral and hyperspectral thermal imagers provide rich spectral information, but high data dimensionality and limited onboard resources make real-time processing challenging. As wildfires increase in frequency and severity, the need for low-latency and computationally efficient onboard detection methods is critical.
  We present a systematic evaluation of multiple deep learning architectures, including custom Convolutional Neural Networks (CNNs) and Transformer-based models, for multi-class fire classification. We also introduce PyroFocus, a two-stage pipeline that performs fire classification followed by fire radiative power (FRP) regression or segmentation to reduce inference time and computational cost for onboard deployment. Using data from NASA's MODIS/ASTER Airborne Simulator (MASTER), which is similar to a next-generation fire detection sensor, we compare accuracy, inference latency, and resource efficiency.
  Experimental results show that the proposed two-stage pipeline achieves strong trade-offs between speed and accuracy, demonstrating significant potential for real-time edge deployment in future wildfire monitoring missions.

تم تقديم PyroFocus كحل يعتمد على التعلم العميق للكشف عن حرائق الغابات في الوقت الحقيقي باستخدام صور الاستشعار عن بعد متعددة الأطياف. تهدف هذه الطريقة المبتكرة إلى تحسين دقة وسرعة تحديد ظروف الحريق، وهو أمر حاسم للاستجابة الفعالة للطوارئ وإدارة البيئة.

PyroFocus se ha presentado como una solución de aprendizaje profundo para la detección en tiempo real de incendios forestales utilizando imágenes de teledetección multiespectral. Este enfoque innovador tiene como objetivo mejorar la precisión y la velocidad en la identificación de condiciones de incendio, lo cual es crucial para una respuesta de emergencia efectiva y la gestión ambiental.

PyroFocus a été introduit comme une solution d'apprentissage profond pour la détection en temps réel des incendies de forêt à l'aide d'images de télédétection multispectrales. Cette approche innovante vise à améliorer l'exactitude et la rapidité d'identification des conditions d'incendie, ce qui est crucial pour une réponse d'urgence efficace et la gestion environnementale.

PyroFocus has been introduced as a deep learning solution for real-time wildfire detection using multispectral remote sensing imagery. This innovative approach aims to enhance the accuracy and speed of identifying fire conditions, which is critical for effective emergency response and environmental management.

PyroFocus: A Deep Learning Approach to Real-Time Wildfire Detection in Multispectral Remote Sensing Imagery

arXiv:2512.03911v1 Announce Type: cross 
Abstract: We present an end-to-end pipeline for deploying reinforcement learning (RL) trained Artificial Neural Networks (ANNs) on neuromorphic hardware by converting them into spiking Sigma-Delta Neural Networks (SDNNs). We demonstrate that an ANN policy trained entirely in simulation can be transformed into an SDNN compatible with Intel's Loihi 2 architecture, enabling low-latency and energy-efficient inference. As a test case, we use an RL policy for controlling the Astrobee free-flying robot, similar to a previously hardware in space-validated controller. The policy, trained with Rectified Linear Units (ReLUs), is converted to an SDNN and deployed on Intel's Loihi 2, then evaluated in NVIDIA's Omniverse Isaac Lab simulation environment for closed-loop control of Astrobee's motion. We compare execution performance between GPU and Loihi 2. The results highlight the feasibility of using neuromorphic platforms for robotic control and establish a pathway toward energy-efficient, real-time neuromorphic computation in future space and terrestrial robotics applications.

تم تطوير خط أنابيب شامل لنشر الشبكات العصبية الاصطناعية (ANN) المدربة بواسطة التعلم المعزز (RL) على الأجهزة العصبية Loihi 2 من إنتل، من خلال تحويلها إلى شبكات عصبية من نوع Sigma-Delta (SDNN). تم اختبار هذه الابتكار باستخدام سياسة RL للتحكم في الروبوت الطائر Astrobee، مما أظهر استنتاجًا منخفض الكمون وفعالًا من حيث الطاقة في بيئة المحاكاة Omniverse Isaac من NVIDIA.

Se ha desarrollado un pipeline de extremo a extremo para implementar redes neuronales artificiales (ANN) entrenadas mediante aprendizaje por refuerzo (RL) en el hardware neuromórfico Loihi 2 de Intel, convirtiéndolas en redes neuronales Sigma-Delta de picos (SDNN). Esta innovación se probó utilizando una política de RL para controlar el robot volador Astrobee, demostrando inferencia de baja latencia y eficiencia energética en el entorno de simulación Omniverse Isaac de NVIDIA.

Un pipeline de bout en bout a été développé pour déployer des réseaux de neurones artificiels (ANN) entraînés par apprentissage par renforcement (RL) sur le matériel neuromorphique Loihi 2 d'Intel, en les convertissant en réseaux de neurones Sigma-Delta à impulsion (SDNN). Cette innovation a été testée en utilisant une politique RL pour contrôler le robot volant Astrobee, démontrant une inférence à faible latence et économe en énergie dans l'environnement de simulation Omniverse Isaac de NVIDIA.

An end-to-end pipeline has been developed for deploying reinforcement learning (RL) trained Artificial Neural Networks (ANNs) on Intel's Loihi 2 neuromorphic hardware, converting them into spiking Sigma-Delta Neural Networks (SDNNs). This innovation was tested using an RL policy for controlling the Astrobee free-flying robot, demonstrating low-latency and energy-efficient inference in NVIDIA's Omniverse Isaac Lab simulation environment.

Autonomous Planning In-space Assembly Reinforcement-learning free-flYer (APIARY) International Space Station Astrobee Testing

Was this article worth reading? Share it

Airparser

Boost.space

Dyad