The growing complexity of space operations and the increasing density of resident space objects (RSOs) in Earth orbit demand highly accurate and timely orbit determination capabilities. Space-based surveillance has emerged as a critical complement to ground-based systems, enabling improved coverage and responsiveness for Space Situational Awareness (SSA). Nevertheless, space-based operational scenarios often involve extremely short observation windows resulting in “too-short arcs” that challenge traditional orbit determination methods. In fact, such constraints lead to high uncertainty in state estimation, sensitivity to initial conditions, and increased computational complexity. Extraction of accurate and reliable orbital states and covariances from sparse, space-based measurements is therefore essential for timely collision avoidance, catalog maintenance, and space operations safety.

This work presents the design and implementation of a robust orbit determination pipeline at ARCA Dynamics, discussing its workflow and inherent challenges associated with short-arc data. After reviewing the main techniques available in the literature, the paper describes the development of an operational pipeline that integrates observation scheduling, image acquisition and processing, orbit estimation, and validation against independent measurements. The performance of such a pipeline can be compared to ground-based observation processing, and its algorithms and mathematical formulations can be validated against independent measurements and simulated scenarios. Additional capabilities, including attitude determination from images and chained estimation techniques, are evaluated to enhance robustness under sparse data conditions. The pipeline is fully operational and has been applied to real-world scenarios, with representative examples included in this work.

Each stage of the pipeline faces limitations, some unique to space-based operations. Scheduling depends on observer-target geometry, illumination, and operational constraints. Image processing must handle noise and pointing precision. Orbit determination is challenged by limited geometric diversity, range ambiguity and initial guess sensitivity. Unlike ground-based observations, the observer’s state and attitude knowledge critically affect accuracy, particularly when third-party data is incomplete or corrupted. These factors collectively amplify uncertainty and complicate real-time processing, requiring robust mitigation strategies.

ARCA Dynamics' experience demonstrates the feasibility of operational space-based orbit determination under severe data constraints. By systematically analyzing limitations and mitigation strategies for short-arc processing, this work contributes to improving SSA capabilities and ensuring safer operations in increasingly congested orbital regimes.