GMV Innovating Solutions (Poland)

ABSTRACT This survey provides a comprehensive overview of recent advancements and challenges in Artificial Intelligence (AI)‐enabled computer vision (CV) techniques for space robotic missions, spanning critical phases such as Entry, Descent, and Landing (EDL), orbital operations, and planetary surface exploration. Emphasis is placed on deep‐learning–based approaches for image classification, object detection, semantic segmentation, relative pose estimation, and feature matching. State‐of‐the‐art methods in terrain‐relative navigation, crater‐based or rock‐feature matching, and pose estimation for uncooperative targets are highlighted, illustrating the progress achieved through hybrid pipelines combining deep neural networks with classical geometry. The paper also critically evaluates publicly available orbital and planetary data sets—along with the increasing role of synthetic data—for developing and benchmarking CV algorithms under strict resource limitations and harsh environmental conditions. Despite demonstrated success in tasks like autonomous landing, debris removal, and rover navigation, current solutions face significant hurdles. These include computational constraints on onboard hardware, insufficient coverage of planetary conditions in existing data sets, and limited adaptability to dynamically changing environments. To address these shortcomings, research must prioritize lightweight neural architectures, advanced synthetic data generation, adaptive or incremental learning, and robust multisensor fusion. By integrating these strategies, AI‐based CV systems can advance autonomy, precision, and resilience in future space missions.

When Words Matter Most: Conveying Serious Health Information to Parents

The chapter presents modification of a dynamically equivalent manipulator (DEM) method, which enables dynamic modeling of space robots and manipulators, including free-floating maneuvers, via their suitable substitution by groundfixed manipulator models. DEM provides an attractive modeling and control design tool, since it enables conducting tests and experiments for space vehicles in earth laboratories. The modification of DEM method relies upon an introduction of quaternion-based modeling of manipulator attitude. Originally, DEM method was developed in Euler angles. The basic motivation for the presented DEM modification is to make dynamic and kinematic models more suitable for description of space robots and manipulator motions and their missions like debris removal, spacecraft servicing, space mining, and on-orbit docking and assemblies. It may also support control designs. The theoretical development is illustrated by an example of generation of spacecraft quaternion-based dynamics and simulation studies of its reorientation maneuvers.

Deep space missions, and in particular missions to asteroids, impose a certain level of autonomy that depends on the mission objectives. If the mission requires the spacecraft to perform close approaches to the target body (the extreme case being a landing scenario), the autonomy level must be increased to guarantee the fast and reactive response which is required in both nominal and contingency operations. The GNC system must be designed in accordance with the required level of autonomy. The GNC system designed and tested in the frame of ESA’s Asteroid Impact Mission (AIM) system studies (Phase A/B1 and Consolidation Phase) is an example of an autonomous GNC system that meets the challenging objectives of AIM. The paper reports the design of such GNC system and its validation through a DDVV plan that includes Model-in-the-Loop and Hardware-in-the-Loop testing. Main focus is the translational navigation, which is able to provide online the relative state estimation with respect to the target body using exclusively cameras as relative navigation sensors. The relative navigation outputs are meant to be used for nominal spacecraft trajectory corrections as well as to estimate the collision risk with the asteroid and, if needed, to command the execution of a collision avoidance manoeuvre to guarantee spacecraft safety

After the success of the NASA rovers (Spirit, Opportunity and Curiosity), the European Space Agency is developing its own mobile platform within the Exomars activity. Next step should be the acquisition and return of soil samples to the Earth by means of mobile robots. Such mobile platforms should provide more payload capabilities, move faster and travel longer than current MER rovers. At the same time they must be able to react against the uncertainty of the always poorly known environment. All these new demanding requirements are leading to the need of autonomy developments for the space and ground systems, driven by the benefits that autonomy brings in terms of reducing mission operational costs, enabling long term missions and maximizing scientific return. In the area of planetary rovers the robotics autonomy must be achieved by implementing autonomous navigation capabilities in the functional layer and autonomous decision-making systems in the deliberative layer. In this paper, we propose an integrated architecture composed by a functional layer complemented with an interleaved deliberative and executive-layer.