Recent research is developing autonomous policies for agile Earth-observing satellite (AEOS) tasking using deep reinforcement learning (DRL). Prior work uses a fixed training environment with enhanced safety challenges to allow DRL to train a policy over a short number of orbits. This paper investigates how training environment enhancements can improve these policies’ robustness by varying the training environment and extending simulated mission times. DRL enables real-time, onboard decision-making while accounting for complex dynamics and constraints. However, policies often struggle to generalize to longer missions or conditions different from training, partly due to the short simulated mission times. Multi-spacecraft systems amplify these challenges, where battery capacity and solar panel efficiency diverge over time due to external factors. Although satellites might run copies of a single policy, its robustness is essential to accommodate agent variability and ensure mission success. This paper investigates training with two curricula, constantly degraded environments, and a domain randomization approach to improve AEOS policy robustness and performance. Policies are tested on episodes 125-times longer than training episodes under nominal and degraded conditions. The proposed training environment enhancements’ applicability is demonstrated in a heterogeneous spacecraft system, achieving a 63% reduction in failures and a 2.8% increase in median cumulative reward.
2 September 2022 in Brazilian Journal of Geophysics
The INPE-UFSM's NANOSATC-BR CubeSats Development Program started in early 2008. Nowadays, the program has two CubeSats in space: the NANOSATC-BR1 (1U), launched in 2014 and currently in standby mode, and the NANOSATC-BR2 (2U), which was recently launched in March 22, 2021 from the Baikonur Cosmodrome, Kazakhstan. In this article, the launch of NANOSATC-BR2 is presented, alongside with its final tests, initial operations and expected scientific results. Regarding the Capacity Building, the main objective of the program, this work focuses in the involvement of the Federal University of Santa Maria - UFSM undergraduate students and professionals in the tracking and control of CubeSats through the INPE’s Ground Station at Santa Maria, which works as an excellent tool for supporting CubeSat scientific missions from all around the world and also to qualify students to work on the Ground Segment of space missions. The program was supported by the Brazilian Space Agency (AEB) and the Ministry of Science, Technology and Innovation (MCTI).
1 January 2022 in Aerospace Science and Technology
The use of unmanned aircraft vehicles (UAVs) for the most diverse aeronautic missions is a reality. Then, with the rise of precision agriculture (PA), the viability of the incorporation of this technology in the countryside on both big and small proprieties is under study. This work aims to analyze the handling qualities of an UAV with a mission of pest control chemical spraying. With this objective, aircraft data was used to analyze the eigenvalues of the aircraft modes. Then, the impact of mass and inertia variation during its mission was studied. Also, the longitudinal and latero-directional dynamics were evaluated based on the Cooper-Harper scale. The results found indicate that all aircraft modes are stable for both full hopper and empty hopper conditions. Moreover, the determined handling quality grades of the UAV in the Cooper-Harper scale indicate the need for some correction applied by the pilot during flight.
1 January 2020 in Journal of Aerospace Technology and Management
During the last decade, the world faced the mass insertion of small satellites in the space technology scenario. Every year, the number of micro and nanosatellites launched increases and gets more attention from players in the space market. Despite the lack of a national launcher, the Brazilian Space Program is known for some successful development in the last century, including its space assets, such as a privileged launch site near the equator, a family of flight proven and reliable sounding rockets for suborbital flights and microgravity experiments and universities with established small satellites programs. Thereby, the present work proposes a modification of the Brazilian VSB-30 sounding rocket in order to allow the launch and insertion in low Earth orbit (LEO) of small satellites fulfilling the gap of a national launcher. It also presents a CubeSat orbital decay simulation and orbital insertion simulation with the modified rocket launched from the Alcantara Launch Center as a matter of verifying the potential of national missions using this modified launcher.
1 November 2026 in Journal of Aerospace Information Systems
This paper investigates the impact of shields on the performance of deep reinforcement learning (DRL)-based policies in the context of autonomous agile Earth-observing satellites. The scheduling phase during spacecraft operations determines the sequence of actions that the satellite should take to meet the mission requirements while adhering to the system’s constraints. The increasing number of satellites, the growing demand for their services, and the need for fast replanning due to unexpected events or additional requests push the need for methods that can provide solutions in real time. DRL has shown promise for onboard scheduling, but deploying neural networks in critical systems, such as satellites, raises reliability concerns. Research has proposed using shielded neural networks (SNNs) to provide safety guarantees with DRL. This work provides a comprehensive analysis of the impact of different approaches to training and deploying SNNs on agent performance. Results show that training with action replacement while assigning penalties for shield intervention leads to higher cumulative reward and lower shield interference during testing. Further, shielded agents performed similarly to the best unshielded agents but with fewer safety violations. Although shields provided safety in 30-times-longer-than-training episodes, policies trained with safety concerns outperform those trained without them.