R. Purschke’s research while affiliated with Technical University of Munich and other places

Orbital robotics is receiving growing attention worldwide for applications in servicing and repositioning of partially or fully defective satellites. In this paper, we present the scope and main results of a four-year research project, which aimed at developing necessary robotic technologies for such applications. The scope is two-fold, since we address both the human-operated robotic operational mode, referred to in robotics as force-feedback teleoperation, as well as the alternative autonomous mode, for the specific task of approaching and grasping a free-tumbling target satellite. We present methodological developments and experimental as well as numerical validations in the fields of tele-communications, computer vision, robot and spacecraft control and system identification. The results of this work constitute important advances in the fundamental building blocks necessary for the orbital applications of interest.

The goal of this work was to (1) evaluate different gear concept for spaceflight application with minimum backlash, (2) design a test setup to test the most promising concepts, and (3) conduct preliminary validation tests. The evaluation focused on the primary possibilities to influence the behavior of gear performance: variation of material and variation of geometry. The synthetic materials PEEK and Vespel were evaluated with respect to vacuum compatibility, wear and strength. Both materials performed promising for space application. To investigate the influence of gear geometry, a profile shift of the gears and a reduction of the tooth height were analyzed and tested. The new gear concepts were investigated in a specially design test setup. It allows the measurement of backlash with an adjustable distance of the two gear axes. The test setup also allows for the testing of friction and wear, which is not part of the presented work. The design of the test bed also allows testing in thermal-vacuum environment. First tests were conducted to verify the functionality of the test setup. Preliminary measurements with two different gear geometries were performed. A larger follow-on study will then test various gear design concepts in thermal-vacuum conditions and has the goal to compare the design solutions regarding application in high accuracy space mechanisms.

The goal of this work was to (1) define parameters to characterize a pointing mechanism, (2) design a setup to test these parameters and, (3) verify the test methods by comparing the results to the theoretically calculated or independently verified numbers. The verification of the test results was conducted with an in-house built Antenna Pointing Mechanism for on-orbit servicing applications. The test setup was developed to find a method to measure the behavior of a pointing mechanism. This was realized by mounting a Laser pointer on the antenna interface of the mechanism and pointing it towards a two-dimensional Position Sensitive Detector, providing means to resolve small motions, and to derive velocity and acceleration of the mechanism. The results show good correlation for characteristic parameters such as pointing velocity and acceleration, repeatability, resolution and pointing accuracy of the mechanism. In future work this test method will be qualified for and used to compare the performance of the mechanism at different environmental conditions such as vacuum, temperature and microgravity.

Teleoperation of spacecraft proximity operations and docking requires delicate timing and coordination of spacecraft maneuvers. Experience has shown that human operators show large performance fluctuations in these areas, which are a major factor to be addressed in operator training. In order to allow the quantification of the impact of these human fluctuations on control system performance and the human perception of this performance, a learning curve study was conducted with teleoperated final approach and docking scenarios. Over a period of ten experiment days, three test participants were tasked with repeatedly completing a set of three training scenarios. The scenarios were designed to contain different combinations of the major elements of any final approach and docking situation, and to feature an increasing difficulty level. The individual difficulty levels for the three operators furthermore differed in the level of operator support functions available in their human-machine interfaces. Operator performance in the test scenarios were evaluated in the fields approach success and precision, docking safety, and approach efficiency by a combination of recorded maneuver data and questionnaires. The results show that operator experience and the associated learning curves increase operator performance substantially, regardless of the support system used. The paper also shows that the fluctuations in operator performance and self-perception are substantial between as well as within experiment days, and must be reckoned with in teleoperation system design and mission planning.

The objectives of on-orbit servicing (OOS) missions include manipulation, proximity operations and inspection of target satellites. Therefore the servicer satellite often has to be teleoperated at low latency for several minutes to fulfill these tasks. That means communication plays a crucial role for OOS missions because real time teleoperation including high data rates has to be realized. So the communication path from front end sensors on the servicer spacecraft to the operator on ground has to be optimized and the latency time has to be minimized. Furthermore a long access time from the ground station is required because continuous communication with the satellite is mandatory for most of the OOS tasks. This can be realized by an inter-satellite link via a geostationary relay satellite, which has the advantage that a satellite in Low Earth Orbit (LEO) can be accessed from one ground station for about half an orbit. To evaluate both, the requirement of a long access time from the ground station as well as the need of a short latency time, an end to end communication scenario was implemented at the Institute of Astronautics (LRT) at the Technische Universität München (TUM). This scenario includes different spacecraft sensors (e.g. stereo cameras, LIDAR systems), a Ka-Band ground station and man machine interfaces. This paper describes the setup of a realistic simulation of a communication path from a data source to an operator via space-link. Furthermore the method of latency measuring depending on the data source is described. The communication architecture is embedded in a spacecraft simulator to simulate On-Orbit Servicing scenarios like Space Debris removal and target inspection.

Since 2008 the lightweight intersatellitelink antenna (LISA MS ) has been developed for establishing communication between low flying satellites (e.g. earth observation satellites) and geostationary relay satellites. For this development a detailed electrical design has been simulated to predict performance parameters. The paper describes the detailed component design and the comparison of the simulated values with measurements of the manufactured breadboard antenna model.