Technologies for Mars On-Orbit Robotic Sample Capture and Transfer Concept ∗ Rudranarayan Mukherjee Neil Abcouwer Jet Propulsion Laboratory Jet Propulsion Laboratory California Institute of Technology California Institute of Technology 4800 Oak Grove Dr. 4800 Oak Grove Dr. Pasadena, CA 91109 Pasadena, CA 91109 Rudranarayan.M.Mukherjee@jpl.nasa.gov Neil.Abcouwer@jpl.nasa.gov (818) 354-2677 (626) 755-5821 Junggon Kim Ryan Mccormick Jet Propulsion Laboratory Jet Propulsion Laboratory California Institute of Technology California Institute of Technology 4800 Oak Grove Dr. 4800 Oak Grove Dr. Pasadena, CA 91109 Pasadena, CA 91109 Junggon.Kim@jpl.nasa.gov Ryan.L.Mccormick@jpl.nasa.gov (818) 393-4146 (818) 354-5945 Peter Godart Philip Bailey Jet Propulsion Laboratory Jet Propulsion Laboratory California Institute of Technology California Institute of Technology 4800 Oak Grove Dr. 4800 Oak Grove Dr. Pasadena, CA 91109 Pasadena, CA 91109 Peter.T.Godart@jpl.nasa.gov Philip.Bailey@jpl.nasa.gov (818) 354-5017 626) 460-9653 Abstract—Potential Mars Sample Return (MSR) would need a robotic autonomous Orbital Sample (OS) capture and manipu- lation toward returning the samples to Earth. The OS would be in Martian orbit where a sample capture orbiter could find it and rendezvous with it. The orbiter would capture the OS, manipulate it to a preferential orientation for the samples, tran- sition it through steps required to break-the-chain with Mars, stowing it in a containment vessel or an Earth Entry Vehicle and providing a redundant containment to the OS (e.g., by closing and sealing the lid of the EEV). In this paper, we discuss component technologies developed for in-laboratory evaluation and maturation of concepts toward the robotic capture and manipulation of an Orbital Sample. We discuss techniques for simulating 0-g dynamics of a spherical OS, including contact, in a laboratory setting. In this, we leverage a 5dof gantry system and, alternately, a 6dof KUKA robotic arm to simulate the OS motion. Both the gantry and robotic arm are mounted with a force-torque sensor that enable detection of contact and provide measurements to simulate, in hardware, the 0-g OS dynamics. We present results that demonstrate the validity of our approach and the extent to which we are able to simulate 0-g dynamics in a laboratory setting. We also discuss techniques for detecting and tracking the OS using optical sensors and LIDAR from near-capture distances. These are discussed in the context of individual sensors as well as fusion of multiple sensor readings. Results of hardware experiments with different sensors are presented. Further, we discuss an uncertainty quantification based physics modeling capability for quantitative evaluation of different concepts for OS capture and manipulation. The computational models are based on high-fidelity multibody dy- namics simulations of the OS, robotic elements and their contact mechanics. We present results that demonstrate our effective use of computational simulations in a complementary manner to hardware experiments. Finally, we present a cyber-physical approach to concurrently fusing hardware elements, computa- tional simulation elements and autonomy software to effectively and rapidly simulate end-to-end systems concepts for end-to-end orbital sample capture and manipulation system concepts. TABLE OF CONTENTS 1. I NTRODUCTION ...................................... 1 2. OS DYNAMICS TEST BEDS .......................... 2 3. OS TRACKING ...................................... 2 4. CAPTURE TEST BED ................................ 4 5. AUTONOMY SOFTWARE ............................. 7 6. COMPUTER SIMULATIONS .......................... 7 7. CYBER-PHYSICAL SIMULATIONS ................... 9 8. FUTURE WORK .................................... 10 9. CONCLUSIONS ..................................... 10 ACKNOWLEDGMENTS ................................ 10 REFERENCES ......................................... 10 BIOGRAPHY .......................................... 11 1. I NTRODUCTION With the success of the Mars rover missions, interest has grown again in formulating a series of potential missions that would acquire samples from Mars and return them to Earth. These are commonly referred to as potential Mars Sample Return (MSR) [1]-[2]. Mars 2020 will take samples, seal them in tubes and leave them on the Martian surface. Concepts currently being considered involve another rover acquiring these tubes, storing them in a cache and loading the cache on a Mars Ascent Vehicle (MAV) in an Orbital Sample (OS) cache. The MAV could then launch the OS into Martian orbit where a subsequent orbiter would rendezvous with it and robotically capture it. Along with the payload for rendezvous and capture, the orbiter would also have robotic manipulation capabilities to orient the OS, within bounds, of a predetermined orientation, take the OS through various plan- etary protection steps (commonly referred to as Breaking the chain or BTC with Mars), stow it in an Earth Entry Vehicle 1 * Pre-decisional: for information and discussion purposes only