Recent studies of the space debris population in Low Earth Orbit (LEO) have concluded that certain regions have reached a critical density of objects, which will eventually lead to a cascading process called the Kessler syndrome, even with the full implementation of the current mitigation measures. Thus, there is a consensus among researchers that the active debris removal (ADR) should be performed in the near future if we are to preserve the space environment for future generations.

Among the proposed ADR capture technologies, those involving orbital robotics are at the moment the most mature ones since they have been successfully tested in-orbit in more than one occasion. However, no robotic spacecraft has ever performed a capture of a completely non-cooperative, tumbling object and not every target can be tackled by this type of method. Therefore, every target needs to be associated with a specific ADR method based on its type, orbital state, attitude state, age, etc. Performing this task is very difficult and time consuming mainly due to the dimensions of the parameter space describing each method and target object. Moreover, currently there is no easy way to express the degree of hazard that an object represents for an ADR mission that could be easily grasped also by a less technical audience, such as the decision makers.

To overcome these limitations, my research aims at developing a rigorous method for classifying LEO space debris objects and associating an ADR method to a specific target, based on its class and hazard metric. Moreover, my research aims at developing a trajectory generation method for the reach phase of the robotic spacecraft that would ease the stabilization phase of the capture of a non-cooperative, tumbling target.

This presentation, in particular, will illustrate the refinement of the hazard metric of the second layer of the envisioned taxonomy. More in detail, refinement of the main characteristics defining levels of non-cooperativeness of objects will be presented along with the more rigorous redefinition of the break-up risk index of objects based on the severity number of the worst possible failure mode and its the probability of occurrence based on the statistical data collected from publicly accessible data.

Moreover, the presentation will also illustrate the 2D implementation of the trajectory generation method and initial results. More in detail, the method has been formalized as a constrained nonlinear optimization problem reduced via a direct method into a nonlinear programming problem (NLP) and solved using the MATLAB’s optimization toolbox. The method has been implemented using the open source OptimTraj MATLAB library which incorporates different transcribing methods, e.g., Hermite-Simpson, Chebishev-Lobatto, Runge-Kutta and provides an easy method for comparing them. The optimization is performed in two iterations meant to provide at first a reasonable guess and a more refined solution.