ROBUSSTATT reaches mid-term progress on robust attitude uncertainty estimation for debris removal

GMV and the University of Strathclyde have reached the midterm milestone of ROBUSSTATT, an ongoing European Space Agency (ESA) activity focused on robust attitude determination and uncertainty estimation for resident space objects. The consortium has researched, developed, and implemented pre-prototype attitude determination methods in six observation types, and these methods are now undergoing testing to identify the approach to be implemented in the final prototype.

Accurate knowledge of a non-cooperative target’s tumbling state is a key enabler for Active Debris Removal (ADR) and In Orbit Servicing (IOS) missions. Operators must make decisions based on incomplete observations, and different sensors provide different pieces of evidence with varying error characteristics, making uncertainty handling as important as the attitude estimate itself.

ROBUSSTATT builds on the ESA’s in-Orbit Tumbling Analysis tool (iOTA), which supports orbit and attitude propagation and can generate simulated measurements for comparison against real observations. The project extends and improves measurement modelling so that robust attitude estimation methods can be exercised consistently in observation types.

The six observation types addressed include passive optical light curves, high-rate photon counting data, hyperspectral or multicolor light curves, inverse synthetic aperture radar (ISAR) imaging, laser ranging residuals, and radar cross-section (RCS) time series. The work connects these measurements to physical and geometrical representations of the target, including modeling choices that support realistic signature generation, as well as computational efficiency measures needed for repeated estimation and testing cycles.

The central objective of the project is to fuse information extracted from multiple observation types using an Epistemic Gaussian Mixture Filter approach, enabling the prototype to provide an attitude state captured in a cumulative distribution function (CDF) that is traceable to the contributing measurements. The project also aligns with established exchange formats, including CCSDS Tracking Data Message (TDM) inputs, and represents the CDF data though graphical methods that support integration into broader analysis and operational toolchains.

During the next phase of ROBUSSTATT, the consortium will translate the pre-prototype code into modular Python libraries to improve reuse and extensibility. It will also complete validation of the attitude determination methods using representative simulated datasets and iOTA, drawing on both historical and new observation data. GMV leads the activity as prime contractor, with responsibility for systems engineering, prototype translation, and performance assessment, while the University of Strathclyde contributes specialist research, pre-prototype method development, and support for measurement simulation enhancements and validation.

GMV develops mission-critical space software and ground segment systems for customers worldwide, with capabilities spanning Space Situational Awareness (SSA), Space Surveillance and Tracking (SST), and flight dynamics. GMV teams combine modeling, estimation, and operational engineering to deliver tools that support safer and more sustainable space operations.