REX-J, Robot Experiment on the ISS/JEM to demonstrate the Astrobot’s locomotion capability

Mitsushige Oda1), Masahiro Yoshii1), Hiroki Kato1), Atsushi Ueta1), Satoshi Suzuki2), Yusuke Hagiwara3), Taihei Ueno3)

1) Aerospace Research and Development Directorate, Japan Aerospace Exploration Agency, Ibaraki, Japan 2) AES Co. Ltd.., Ibaraki, Japan 3) Department of Mechanical and Aerospace Engineering, Tokyo Institute of Technology, Tokyo, Japan

A unique space robot named Astrobot ( + Robot) is being developed and will soon be demonstrated on the International Space Station, Japanese experiment module KIBO. Tasks of the robot will include supporting ’ works in space and working instead of astronauts. To work with or instead of an astronaut, the robot needs to be able to moves around / inside the space facility, e.g. a space station and need to conduct tasks like an astronauts. The Astrobot’s locomotion capability is realized by an extendable robot arm and tethers. Tethers will be anchored to a handrail or other suitable anchoring points using an extendable robot arm. This unique mechanism of the proposed robot makes it possible to realize the robot in a small volume while the robot can move around the wide area. In order to demonstrate usefulness of this unique robot, an onboard experiment on the exposed facility of the International Space Station Japanese Experiment Module, “KIBO” will be conducted in the year 2012. Development of the experiment system is progressing now.

Key Words: International space station, Astronaut, Space robot, STEM, Astrobot

1. Introduction exploring the moon and the planetary surfaces. Construction of the International Space Station However those robots are not good at helping Japanese Experiment Module, KIBO is now completed astronauts’ IVA (Intra vehicular activities) and and full utilization of KIBO for development of new EVA (Extra vehicular activities). Therefore, technologies for future space missions and improving space robots of a new type are necessary to reduce knowledge and life of people on the . JAXA is now the workload of astronauts. This paper presents a implementing various missions to utilize KIBO. JAXA is new type of space robot, Astronaut Support-Robot to conduct a space robot experiment on the international (Astrobot), that can move inside and outside the space station Japanese experiment module (ISS/JEM). space station and which can conduct tasks that The experiment is named as REXJ (Robot Experiment on are currently conducted by astronauts. JEM). Aim of the REXJ experiment is to demonstrate capability of a new type of space robot named Astrobot (Astronaut + Robot). There are many tasks on the space station to be conducted by astronauts. Some tasks are simple but need many hours. Some tasks are dangerous to conducts. Some tasks need extensive skills. However, number of astronauts onboard the space station is limited. Therefore, from interests of safety and economy, some tasks should be conducted by robots and some tasks should be conducted by astronauts. Robots are also welcomed to assist astronauts while they conduct some tasks. The Astrobot is a type of space robot that will support the astronaut by assisting astronaut’s work or conducting works instead of astronauts. The Astrobot is a new type space robot beyond the orbital robots and the lunar / Fig.1. Astronaut conducting Extra-Vehicular Activity planetary exploration robots. The remote manipulator system and the space station’s remote 2. Tasks of astrobot manipulator system are typical examples of the orbital Interviews with astronauts were conducted to robots. The finder and the Mars Exploration Rovers determine the actual needs and requirements for are typical examples of the exploration robots. the Astrobot. Those interviews revealed the The above orbital robots and the exploration following. robots are good at handling massive payload and x Astronauts are eager to accomplish assigned

1

i-SAIRAS 2010 August 29-September 1, 2010, Sapporo, Japan 567 missions. to move around the space facilities such as the x Astronauts are busy in space. For that reason, space station. Each method has advantage and they want to use their limited time in space disadvantages. efficiently. x Astronauts are burdened with many tasks. Method Example Advantage / Some tasks, such as EVA, are very important; Disadvantage astronauts are eager to accomplish them. Free Flying AERCam Risk of collision, limited x Many preparatory tasks must be accomplished, life by the fuel consumption e.g. transporting equipment to the EVA work Attached to a Dextre Limited locomotion area site to do EVA tasks such as installing large RMS by the reach of RMS equipment or repairing the facility Move on rails SSRMS Locomotion area is x Some tasks such as monitoring equipment and limited by the reach of facility inspection are time-consuming, RMS and rails although they are simple to conduct. Inchworm SSRMS Locomotion area is motion using limited by the location These findings suggest that a support-robot grapple of the grapple fixture should be designed to conduct at least the fixture following tasks. Multi arms EUROBOT, System becomes x Simple but time-consuming tasks: Such tasks robot that can Robonaut complex since number of grasps degrees of freedom include monitoring equipment, inspection of handrails increase. outer surfaces of the space station (inspection Move by Charlotte Size of robot is small of impact damage by debris), and monitoring tethers while locomotion area is of astronauts’ activities. wide x Transporting equipment: conveying and Table 1. Ways of robots’ receiving equipment to and from the astronaut. In the future, robot(s) will also build and Since most of locomotion methods have limited maintain large space facilities such as the locomotion area while their systems are solar power . To build the solar power complicated, we decided to use tethers to move a satellite economically, robots, not astronauts, robot around the space station. Tethers will be should be used to build the facility. connected to handrails that are originally prepared to be used by astronauts to support their 3. REXJ mission body. On the international space station, there are REXJ is an acronym of the Robot Experiment on many handrails are attached at an interval of one the Japanese experiment module of the ISS. meter or so. Astronaut grasps this handrail when they work outside the space station. (See Fig,1.) 3.1 Objective of REXJ mission Purpose of the REXJ is to demonstrate some key 3.3 Astrobots’ locomotion principle technologies to realize the Astrobot. Technologies Since already proposed space robot’s locomotion to be demonstrated by the REXJ mission include methods have critical disadvantage, we decided to following capabilities. use the tethers to move the robot around the space x Astrobot’s capability to move around the station as follows. Figure 2 depicts the principle of space station and other space structures the robot’s locomotion; using infrastructure prepared for the astronaut’s work. x Astrobot’s capability to manipulate equipment or tools designed for astronauts.

3.2 Astrobot’s locomotion capability Astrobots must be able to move around the space structures such as the space station like an astronaut. If a robot needs some special equipment on the space station and those are not included in the current space station, then such special equipments will not be attached unless those are highly required. Then the robot cannot Fig. 2 Principle of Robot Locomotion move around. There are several ways of locomotion as follows (1) The robot has several tethers inside the robot 2

568 body. Tethers are wound in reels. Each tether Most robot hands developed for on-ground has a hook-like mechanism to attach the commercial applications do not satisfy this requirement. tether to a structure, such as a handrail, They have very limited grasping power because of the which is prepared for astronauts. actuators’ capability to drive fingers. Most hands include (2) Robots have an extendable robot arm. The rotary actuators in the finger joints, but humans’ hands robot arm has a robot hand at its end. are too small to contain sufficiently powerful actuators. (3) The extendable robot arm will grasp the tether Most powerful robot hands are gripper-type robot hands, but those hands are insufficiently dexterous to do many hook and extend the tether. 4) (4) It attaches the tether hook to a handrail or tasks. Two famous robot hands, the Stanford-JPL hand and the Utah/MIT hand5) developed in the 1980s, had secures itself by some other method. large actuators inside of the robot’s body to increase (5) Retract the robot arm and grasp the other grasping power, but the actuators made the hand tether hook. impossible to use depending on the situation and payload. (6) Connect other tethers to other points. Furthermore, this method complicates mechanisms and (7) Adjust the length of each tether. Then the maintenance. location of the robot will change. These analyses indicate that a robot hand for (8) The area in which the robot can move depends Astrobots should have the following functions. on the number of tethers attached to the (a) The robot hand must have sufficient dexterity to structure and the location of each tether grasp or handle payloads with high grasping power; anchoring point. Using three tethers, the area it should be small to be installed in many arms and in which the robot can move is a triangular must be able to handle objects a human could usually plane made by three tether-anchoring points. handle. The area in which the robot can move becomes (b) The robot hand should be removable or exchangeable a three-dimensional space if the number of from its wrist so that it can be maintained easily or tether-anchoring points is four or greater. exchanged depending on the mission. (9) If necessary, change locations of the tethers’ A team of JAXA, THK, KEIO University, and the hooks using the extendable robot arm. Then Tokyo Institute of Technology is developing a robot hand the area or space in which the robot can move that will have the grasping power and dexterity of an can be changed. astronaut in an EVA environment. The hand will have several fingers to grasp and operate a Pistol Grip Tool. This method presents many advantages in comparison Actuators to drive fingers are mounted inside the finger. to the inchworm and flight methods. The robot is All functions necessary to drive the hand such as anchored to the space facility by several tethers. actuators and control electronics must be installed inside Therefore, there is no danger of losing the robot. the hand itself. Therefore, this hand is exchangeable from Fig.3.and 4 show typical examples of Astrobots. A its wrist. similar experiment (Charlotte) 3) was conducted on the space shuttle (STS-63) in 1995. A major difference between the REXJ and the Charlotte is that REXJ robot can decide the area in which the robot will move because tethers are attached to nearby handrails, whereas Charlotte’s tethers must be attached to nearby structures by the onboard astronaut use.

3.4 Astrobots’ manipulation capability The Astrobot must handle (grasp, receive, hand over, manipulate) equipment and tools that astronauts handle. Not only an extendable arm, but also the manipulation arm(s) might be needed to realize all of those functions. Fig.3 JXA-THK hand (FY2010 Model) Because of limited resources, the Astrobot has only a deployable arm and a simple hand to attach the hooks or grab the handrail; future robots might have more rigid arms and more dexterous robot hands. Even if the Astrobot has a more dexterous hand in future missions, it is unrealistic to develop a perfectly multi-purpose robot hand to accommodate both massive and small, lightweight payloads. It is more realistic to exchange several task-oriented robot hands depending on the task. In addition, the robotic hand must have both sufficient grasping force and dexterity to accomplish Fig. 4. JAXA–THK hand (FY2010 model) missions. 3

569 3.5 Astrobots’ applications meaning full missions, an Announcement of If the Astrobots become operational, they can be opportunity (AO) was issued in the year 2006. used in such as following application areas.

(1) Monitor of equipment on the space station (IVA work) (2) Visual Inspection of the space station to identify and investigate damage to the outer surface of the space station. (3) Assemble a large space facility such as a solar power satellite.

Fig. 7. Exposed facility of KIBO

4 Development of the REXJ mission

4.1 Selection of missions REXJ mission was proposed in accordance with the announcement of mission opportunity (AO) on the exposed facility of the Japanese experiment Astrobot module. The AO was issued on November 2006. REXJ was selected as one of the missions to be conducted in the year 2012. The selected missions Fig.5 Astrobot monitoring onboard equipment are as follows. x IMAP (, , upper , and Plasmasphere mapping㧕 x GLIMS (Global and Measurement Mission㧕 x SIMPLE (Space Inflatable Membranes Pioneering Long-term Experiments㧕 x REXJ 㧔Robot Experiment on JEM㧕 x HDTV (High Definition TV)

These missions share one payload unit to be attached to the exposed facility of KIBO as shown by Fig.8 and Fig.9.

Fig.6 Astrobot building the solar power satellite

3.6 In-orbit technology demonstration The International Space Station Japanese experiment module, KIBO has an exposed facility which accommodates payload units as much as 12. Figure 4 shows the exposed facility which accommodates 6 payload units including a pallet to carry up to 3 payload units. Fig.8 A payload unit to be attached to the At the beginning of KIBO’s operation, half of exposed facility of KIBO those ports will be occupied. The remaining ports will be used by later missions. The size of the payload unit is shown in Fig.4. In order to collect

4

570

Fig.9 Artist’s image of the shared payload unit

4.2 Development schedule The above selected missions’ payloads must be built before August 2010 to be integrated into one payload unit. The integrated payload unit will be delivered to the space station by the HTV(H-II transfer vehicle) in the year 2012. The REXJ mission’s development schedule is as follows. x Announcement of opportunity: Nov.2006 x Selection of missions: December 2007 x System Requirement Review: May 2008 Fig. 10. REXJ experimental system on KIBO. x Mission Definition Review: Aug.2007 x System Definition Review(SDR): Oct.2008 4.4 Mission flow and success level x Request for Proposal (RFP): Oct. 2008 Several experiments are to be performed by the x Start of the REXJ phase B: Nov.2008 REXJ mission. The list below shows those to be x Preliminary Design Review (PDR) March 09 conducted on KIBO. x Critical Design Review (CDR); Oct. 2009 Minimum success level experiments: x Delivery of REXJ PFM to the upper system (1) Release of the launch locks and initial Aug.2010 operation checks. x Launch of REXJ; Jan.2012 (2) Extension of the extendable robot arm. Vibration of the extendable robot arm will be 4.3 Experiment system measured when the robot hand and wrist are Because the volume of the space allocated for the moved. REXJ mission is limited and because safety requirements Full success level experiments: are severe, REXJ system has limited function compared (3) The tethered hook is grabbed by the with the to-be-developed operational Astrobot, the REXJ extendable robot arm and then it is attached robot system consists of following subsystems to the handrail. z The robot’s body including the launch lock (4) Extendable robot arm’s vibration property, mechanism positioning accuracy, and the sympathetic z Extendable robot arm and a robot hand at its end vibration with tether and SRA are measured. z Tether systems (A set of a tether reeling mechanism, (5) Releasing and Re-Attachment of the tethered tether and a hook mechanism) hook from/to the handrail by the extendable The robot’s body comprises layered structures. The root arm is demonstrated. advantage of this structure is that it is easy to add (6) Astrobot locomotion using the tethers is functionality merely by adding the extra layer, such as an demonstrated. extra tether layer, a stabilizing leg layer, and manipulation arm layers. For this mission, the Astrobot Extra success level experiments: will have two layers: the extendable robot arm layer and (7) Cooperative control of the extendable robot the tether systems layer. arm and tether mechanism. Figure 10 show an artist’s image of the (8) Extend the extendable robot arm to the experimental robot system and installation in the outside the APU and check the vibration space station’s Japanese experiment module. caused by the change in the temperature and solar irradiation.

Fig.11 shows motion of the REXJ’s robot. 5

571 is 2500 [mm]. The STEM reel mechanism has two sprockets; sprocket holes are on the STEM surface.

Fig. 12 Engineering model of the extendable robot arm (PFM)

Fig.11 REXJ’s robot motion by the tether control. The deployment and retraction mechanism is portrayed in Fig. 11. The test-model has two 5 Major subsystems motors: one actuates the reels of the STEM and is used for roll up; the other actuates the sprocket 5.1 Extendable robot arm and is used for deployment. The sprocket is The key technology for this method of locomotion is actuated and the STEM is pulled out from the the deployable robot arm, which attaches the tethers to STEM reels when the STEM is deployed. The distant handrails. The international space station has STEM stiffness increases as its cross-sectional handrails within a distance that an astronaut can reach shape becomes circular. Immediately after from one to another. For that reason, the robot arm must deployment from the reel, the cross-sectional extend at least more than a human arm’s length. For shape is flat and buckling occurs easily. To prevent greater efficiency, the arm length should be much longer, this, the sprocket is set slightly distant from the but the whole arm should be sufficiently retractable to fit reel so that the tensile stress is applied to the inside the robot. In the REX-J project, the robot size is STEM, whereas its cross-sectional shape is too flat. planned to be less than 500 [mm] × 400 [mm] × 350 The STEM reels are actuated to roll up when it is [mm]. We specifically examine the Storable Tubular retracted. This is also efficient to prevent the Extendible Member (STEM) to realize such high deployment capability. STEMs ease in the reels. Actually, STEM has been known as a simple and reliable deployable structure since the 1960s. However, 5.2 Tether system most STEM applications are antennas and telescopes8) 9) The tether unreeling and reeling mechanism is 10). Therefore, a compressed force is not considered to be rather simple. Figure 12 shows a BBM of the supplied. Most of those STEMs are manufactured by tether reel mechanism. A mechanical hook is Copper-Beryllium, but they are quite heavy to be installed attached at the end of the tether. This hook can be in small or robots. In addition, those former operated by a human or by a robot. STEMs are deployed by their own stored energy; they are unsuitable for deployment under compressed conditions.

Design Concept We are developing an extendable robot arm using the STEM mechanism. The function of the extendable robot arm is to drag and hook/remove tethers to/from the handrails. To satisfy those functions, the extendable robot arm must be: z deployable and retrievable many times stably Fig. 13. Tether reel mechanism and the hook(BBM) without hysteresis. z controllable and able to maintain length, even 5.3 Robot hand under stressed conditions. The Robot Hand segment of the Astrobot consists of a robot hand, wrist joints, and control electronics. As it Mechanism described above, because of the limited resources, only A test model (breadboard model: BBM) of the the simple robot hand can be installed to the REXJ extendable robot arm was produced. A STEM project. However, the robot hand must still grasp at least manufactured using CFRP is used to meet the two different objects: a hook and a handrail. Moreover, in previously described requirements. The Bi-STEM the future, many more objects should be grasped by the method is used to raise the stiffness. The same kind of hand. Therefore, we chose the human-like maximum length of the STEM used by the REXJ multi-joint fingers. The design of the finger follows the 6

572 JAXA-THK hand. The same type of actuator will be used 6. Conclusions in the fingers. JAXA’s research on the astronaut support robot was introduced herein. Following the analysis result of space robots and interviews with astronauts, we proposed a new type of space robot called the Astrobot. This robot has several tethered hooks and a deployable robot arm; it moves using a new locomotion method. Flight experiments on the space station KIBO are anticipated for 2012. The fundamental capability of the Astrobot will be demonstrated. Each component is now under development.

References Fig.14 Tip portion of the extendable robot arm (Wrist joints, hand, monitor cameras) and grasped hook (PFM) 1) Oda, M.: Tethered robot which moves along a large space structure, Proc. Space Sci. Tech. The robot wrist has 2 degrees of freedom, which is the Symp. 2007, Sapporo, Japan. (in Japanese), minimum number to grasp both a vertical and a horizontal 1D03 handrail from any direction. The robot hand unit will be 2) Lovchik, C.S. and Diftler M., A.: The Robonaut Hand: a installed at the top of the STEM. It is impossible to drag dexterous robot hand for space, Proc. IEEE Conf. Robot all cables to drive and control the unit. For this reason, the Automat., (1999) pp. 912–997 control electronics for the all the motors in the hand unit 3) Swaim, P.L, and Thompson. J.T. and Campbell P.D.: The will be installed not in the robot body, but in the wrist. Charlotte, Intra-Vehicular Robot, Proc. Int. Symp. AI, Robot. Autom. Space (i-SAIRAS’97), Jpn, 1996. pp. Design concepts for the robot hand unit are listed 157–162 below. 4) Mason M.T. and Salisbury J.K.: Robot Hands and the Design Concept Mechanics of Manipulation, MIT Press, Cambridge MA, (a) The grasping targets are both the handrail and the 1985. hook of the experiment system. 5) Jacobson S.C. and Iverson E.K.: Designe of the (b) The mechanism and parts are expected to be similar Utah/M.I.T. Dexterous Hand, Proc. IEEE Intl. Conf. to the JAXA-THK hand. Robot. Automation. San Francisco, 1986, pp. 1520–1532. (c) A robot hand has two fingers. 6) Marks G.W.. Reilly M.T, Huff R.L.: The Lightweight (d) The wrist has 2 degrees of freedom. Deployable Antenna for the MARSIS Experiment on the Mars Express , Proc. Of the 36th Aerospace (e) The control electronics should be set in the wrist. Mechanisms Symposium, Glenn Research Center, May 2002. 5.4 Cameras 7) Higuchi K., Watanabe K., Watanabe A., Tsunoda H. and Small CMOS camera will be installed on the robot Yamakawa H.: Design and Evaluation of an Ultra-light hand for inspection and control of the robot. The camera Extendible Mast as an Inflatable Structure, 47th AIAA/ must be sufficiently small enough to be put in the robot ASME/ASCE/AHS/ASC Structures, Structural hand. The REXJ experiment system will have a few more Dynamics, and Materials Conference, Newport, RI. , cameras inside and outside of the robot to monitor the 2006 experiment. 8) Thomson M.W.: Deployable and Retractable Telescoping Tubular Structure Development, The 28th Aerospace Fig.14 shows an artist’s image of the camera’s image Mechanisms Symposium, May 1994, pp. 323–338. (SEE during the onboard experiment. N94-33291 09-15).

Fig. 15. Artist’s Image of the onboard camera Image

7

573