CubeSat Robotic Grasping Mechanism for In-Orbit Manipulation

CubeSat Robotic Grasping Mechanism for In-Orbit Manipulation This paper proposes the design and development of a proof-of-concept Robotic manipulator prototype for various in-orbit tasks such as grasping and interacting with objects. The proposed robotic manipulator features a three-pronged, tendon-driven end-effector. The manipulator’s footprint is contained within a 3U CubeSat while allowing space for other vital subsystems. The current state of the project presents a fully manufactured prototype robotic manipulator with considerations for further functionality. Future work will continue the iterative design process with an investigation of potential new materials and an adaptation for testing the manipulator on a planar microgravity simulator. Nomenclature = Computer-Aided Design = Direct Current = Degree(s) of Freedom = International Space Station = Laser, Imaging, Detection, and Ranging = Linearly Translating Joint = Rotating Joint = Universal Asynchronous Receiver-Transmitter = Universal Serial Bus 3 = 30 cm X 10 cm X 10 cm Cube Satellite 1 = 10 cm X 10 cm X 10 cm Cube Satellite I. Introduction The design process began with a literature review aimed at exploring robotic manipulators. The first design consideration was performed on the types of joints that could be used. For a robotic arm deployed on a small cube satellite, size is a limiting factor. In considering revolute and prismatic joints, it was concluded that the proposed design necessitated a series of revolute joints. For this project, the work done by Baraff, Witkin, and Kass was of particular interest [1]. Following the decision to use revolute joints, these joints’ layout and kinematics were considered [2]. Robotic systems ranging from the Canadarm [3] to aquatic systems [4] were investigated. A key factor for design consideration was complexity. Therefore, a joint layout similar to the OSIRIS-REx (or the aquatic arm pictured in Fig. 1.) was selected. Utilizing revolute joints maintains the translation motion of the arm in two dimensions while allowing pitch and yaw of the end effector. The workspace plane of the arm was positioned to align using attitude control of the CubeSat. Future iterations and improvements to the design and optimization are expected. The final design consideration was the end effector design. End effectors are used for many applications such as painting, welding, and securing items [5–8]. For the mission, a robotic manipulator must be capable of interacting with a wide variety of objects. For this purpose, several different manipulators were considered. A very simple yet viable solution came from the aforementioned OSIRIS-REx which featured a two-pronged gripper. Pronged grippers (and eventually adding joints to resemble fingers) became the chosen interface for the current end effector. The Self Robotic Auxiliary Hand [9] is an under-actuated capture mechanism. The under-actuated prongs can all move independently, which is outside of the scope, but it may be possible to achieve similar results with all prongs actuating at the same time. The proposed robotic arm was inspired by the OSIRIS-REx aquatic manipulator (see Fig. 1). Fig. 1 Minimalist Aquatic Robotic Manipulator [4] Researchers [9, 10] proposed an end-effector to interact and pick up objects, similar to the end-effector found on Astrobee [11–15] onboard the ISS. When designing the prongs of the end effector, consideration was based on the IEEE journal article by S.B. Backus and A.M. Dollar [16]. This article displays a case where the gripper adapts to the grabbed object. The design of the proposed CubeSat manipulator needs to account for multiple points of contact. The proposed method is to utilize three prongs instead of two with a non-stick adhesive, geCKo tape [17], being used. GeCKo tape operates under the principle of the Van der Waals force, capable of withstanding up to 120 kPa in shear stress when normal stresses are induced. To take full advantage of this effect, inspiration was taken from the under-actuated robotic gripper shown in Fig. 2a. This design allows for a pliable grip, increasing contact area with the object of interest. The manipulator proposed in this paper is further optimized by having the third segment of each prong solely spring-actuated. This differs from the first and second segments, which are both tendon-driven and spring-actuated (see Fig. 2b). This design alteration further maximizes surface area contact when manipulating objects, as the third segment of each prong is flexible and preferential for inducing shear forces. (a) Under-Actuated Robotic Gripper [10] (b) Prototype End Effector Prong Fig. 2 Literature Example [10] as compared to Proposed Prototype End Effector Prong II. Proposed Mission The goal of this project is to fabricate, assemble, and adapt a robotic manipulator for a 3U CubeSat mission involving in-orbit object manipulation; however, the current proposed mission focuses on testing the manipulator on a planar microgravity simulator prior to integration. The robotic manipulator operates within a planar workspace, where the first three joints extend and rotate the end effector within the plane. Due to this constrained workspace, aligning the work plane with the simulated microgravity plane will provide insight into the in-plane dynamic response of a space vehicle while operating the arm. Additionally, this test setup will verify the arm’s ability to interact with objects in a microgravity environment. The end effector’s gripping mechanism will incorporate geCKo tape [17] to enhance object security when the three prongs contract. The first iteration of this project aims to develop and test a 4-DOF robotic arm within a 3U CubeSat form factor. The arm will be evaluated based on its ability to operate remotely and collect objects of varying sizes and textures. While future work will integrate the manipulator with a planar microgravity simulator, initial design iterations will focus solely on the robotic arm itself. This includes the development of the motor control system, power systems, end effector, and structural mechanisms. Success will be defined by the manipulator’s ability to accurately grasp and manipulate all test objects. This paper presents the proposed grasping manipulator and details the system’s hardware and software architecture. It includes an overview of the preliminary CAD design, initial hardware fabrication, and assembly process.