Airbus Defence and Space

Space in daily life All features

INVERITAS: Robotics for cleaning up space

Removal of space debris

Space robotics expertise is a core competence of growing importance. INVERITAS is a typical example of the steps Airbus Defence and Space is taking to drive forward research in this field.

INVERITAS (innovative technologies for relative navigation and capture of mobile autonomous systems) is a collaborative research project involving Airbus Defence and Space in Bremen, Jena-Optronik and the Robotics Innovation Centre of the German Research Centre for Artificial Intelligence in Bremen (RIC DFKI). These partners have set themselves the task of developing a prototype multi-mission-capable rendezvous and capture system and the associated core technologies up to technology readiness level 4 (TRL 4) – in other words creating a ground demonstration model.

Model of a tumbling target satellite in the INVERITAS system demonstrator.

Model of a tumbling target satellite in the INVERITAS system demonstrator.

The purpose of the project is to develop technologies for the construction of what are known as servicing robots – spacecraft capable of capturing end-of-life satellites and large pieces of space debris. Once they have been captured, the satellites can either be repaired or guided into a controlled re-entry trajectory. Airbus Defence and Space has allocated part of its own research budget to develop its technological expertise in this specialist area. The German Aerospace Centre (DLR) is providing 50% of the funding.

“Over the past decades, a considerable quantity of human artefacts has been launched into space,” says Airbus Defence and Space project manager Dr Bernd Mädiger. This has led to an accumulation of space debris that is now beginning to threaten future space missions. “Each new collision produces a new shower of debris, greatly increasing the overall number of fragments and thus increasing the probability of damage to passing satellites, a phenomenon known as the Kessler effect.”

Once space debris exceeds a given concentration, it poses an increasing threat to spacecraft in that orbit. Certain orbits are now close to reaching this at-risk situation; hence the urgent need for effective countermeasures. “The risk of collision can be reduced by deorbiting the major structures responsible for producing the largest amount of space debris,” says Bernd. On an annual basis, this would entail between five and ten of the largest objects, weighing between 200kg and 800kg.

A solution based on proven technology

The development of robotic space technologies calls for a high level of engineering skills in a diverse range of disciplines including artificial intelligence, autonomous systems, virtual reality, miniaturisation, materials science, mechatronics, and information and communication technology. In many of these areas, previous experience in related projects such as the development of the Automated Transfer Vehicle (ATV) can serve as a useful starting point. The ATV carries out rendezvous and docking manoeuvres with a ‘cooperative’ target, namely the International Space Station (ISS). The next logical stage, according to Bernd, is to build on this expertise to enable spacecraft to dock with non-cooperative targets.

A cooperative target is capable of identifying itself, is equipped with attitude control systems and visual markers, and is normally able to receive and send radio signals. By contrast, uncooperative targets, such as an inoperative satellite, are passive objects. They emit no signals enabling them to be located, have no identification markers, and in most cases spin around at random because they have no attitude control system to stabilise their orbit.

Dr Bernd Mädiger, INVERITAS project manager at the Airbus Defence and Space robotics laboratory in Bremen.

Dr Bernd Mädiger, INVERITAS project manager at the Airbus Defence and Space robotics laboratory in Bremen.

Bernd expands: “We are nevertheless able to identify them, and we know their model characteristics, but we cannot control them by means of remote commands. We might be able to determine their trajectory, but not their orientation. That is something we have to analyse during the capture mission.”

Venturing into new territory


The technology components contributed by Airbus Defence and Space to the INVERITAS project have called for huge efforts on the part of the design engineers involved. One of the most demanding aspects was the multimodal sensor data processing component, which is required to read, process and merge the complex data delivered by diverse sensor systems and correlate this information in a functional context, including 3D image data. “There is no single sensor that we can use to manage all of the necessary operations, from locating the satellite to defining the approach path,” points out Bernd. This is a task that requires the use of several camera systems, supplemented by a LIDAR laser detection and radar scanner, supplied by Jena-Optronik, and possibly a radar system.

When the data delivered by the sensor systems have been processed, it is the turn of the GNC (guidance, navigation and control) systems to set to work. Their job is to calculate the flight trajectory, for instance when approaching a satellite, and to regulate the distance to the servicing robot once it gets close to its target. Bernd elaborates: “The target in this case is tumbling, i.e. its attitude is constantly changing, and yet we still have to move in extremely close, to a distance of no more than one or two metres, while avoiding all risk of a collision.”

Another of the INVERITAS components is the camera- and LIDAR-based hazard detection and avoidance system for lunar landing. It is so called because the image processing techniques it employs are also being designed with possible future lunar missions in mind. As Bernd explains: “This project has the added bonus that the structure-recognition algorithms we are developing can also be employed in future fully automated landings on the Moon, during the final approach stage, to detect the position of obstacles and initiate the appropriate avoidance measures – because the last thing you want is for the lander to happen to hit a piece of rock.”

The various systems on the servicing robot have to be able to communicate with one another in order to carry out their designated tasks. Image processing, GNC, a robotic arm for grabbing objects, the satellite platform itself – each of these subsystems represents an intelligent unit in its own right. What is needed now is a concept that fuses them together and provides the necessary coordination functions.

The experimental facility housing the INVERITAS system demonstrator on DFKI's premises.

The experimental facility housing the INVERITAS system demonstrator on DFKI's premises.

Practical testing

Project partner DFKI in Bremen has provided access to a large hall measuring 15m by 20m in which two robots are installed. One of them is a small industrial robot capable of carrying weights of up to 60kg, which are being used in an experimental set-up to reproduce the tumbling movement of the target. There is also a large rope and pulley system to which the servicing robot and its satellite platform are attached. A variety of lighting sources are used to illuminate the scene from the side. This set-up enables simulation of the entire rendezvous process.

Is INVERITAS a difficult project? “I would say that we are working on something that borders on the limits of present-day technology. That applies in particular to the sensor data processing component, where the lighting conditions make things extremely difficult,” replies Bernd. “In space, you only have hard shadows: an object is either in complete darkness or extremely brightly lit, depending on its position relative to the sun. There is nothing in between, and the changes take place very rapidly. It takes a highly sophisticated image processing system to analyse the camera data in order to detect contours or individual fragments in space. Another major challenge is that of providing sufficient data processing capacity for the image data to be processed in orbit.”

Up to now, all review milestones have been passed without any problems. Sensor and algorithm tests are currently being carried out in the DFKI simulation facility. INVERITAS is running absolutely to schedule, and will be concluded as planned at the beginning of next year. So, what’s next? “There are other research and technology projects in the pipeline that will build on what we have achieved so far. They will focus on commercial strategies for the removal of space debris.” 

Financed by the German Aerospace Centre (DLR) with funds provided by the German Federal Finance and Technology Ministry under parliamentary decree (project funding reference number 50RA0908).

TechnologyRoboticsSpace surveillance (SSA)