Lunar surface exploration using mobile robots

Cent. Eur. J. Eng. • 2(2) • 2012 • 156-163 DOI: 10.2478/s13531-011-0072-z

Central European Journal of Engineering Lunar surface exploration using mobile robots Vision Paper Shin-Ichiro Nishida∗ , Sachiko Wakabayashi Lunar and Planetary Exploration Program Group, Japan Aerospace Exploration Agency

Received 24 May 2011; accepted 23 February 2012 Abstract: A lunar exploration architecture study is being carried out by space agencies. JAXA is carrying out research and development of a mobile robot (rover) to be deployed on the lunar surface for exploration and outpost construction. The main target areas for outpost construction and lunar exploration are mountainous zones. The moon’s surface is covered by regolith. Achieving a steady traversal of such irregular terrain constitutes the major technical problem for rovers. A newly developed lightweight crawler mechanism can effectively traverse such irregular terrain because of its low contact force with the ground. This fact was determined on the basis of the mass and expected payload of the rover. This paper describes a plan for Japanese lunar surface exploration using mobile robots, and presents the results of testing and analysis needed in their development. This paper also gives an overview of the lunar exploration robot to be deployed in the SELENE follow-on mission, and the composition of its mobility, navigation, and control systems. Keywords: Lunar exploration • Space Robot • Mobility • Remote Control

© Versita sp. z o.o.

1.

Introduction

Recent missions (SMART-1, Kaguya, Chang’E, Chandrayaan-1, LCROSS, and LRO) have been undertaken for active lunar exploration. A lunar orbiter Kaguya (SELENE) was launched in September 2007 by the Japan Aerospace Exploration Agency (JAXA). The spacecraft observed the moon for one and a half years, which yielded favorable results and a considerable amount of scientific data [1, 2]. One of the goals incorporated in JAXA’s long-term vision [3] is to advance the exploration of the moon’s surface; thus, missions to survey and investigate the lunar surface,

∗

E-mail: [email protected]

as a continuation of SELENE’s operation, are being considered. The southern polar region of the moon is the leading candidate location for the construction of a lunar base because of the good sunlight conditions and the high possibility that rocks from the far side of the moon, such as the South Pole Aitoken Basin, are present. The missions are expected to use robotic technologies such as mobile rovers. However, the lunar surface is an adverse environment that presents many technical challenges to exploration and survey activities, particularly since there are many unknowns related to the geographical features and environment in the southern polar region. This paper introduces the strategy of the SELENE followon mission and future lunar exploration. It presents the main technical issues faced by designers of lunar exploration robots, a technical roadmap of their development, and the results of studies of a robot system configuration.

156 Unauthenticated Download Date | 9/24/15 11:36 PM

missions are expected to use robotic technologies such as mobile rovers. However, the lunar surface is an adverse environment that presents many technical challenges to S.-I. Nishida,and S. Wakabayashi exploration survey activities, particularly since there are many unknowns related to the geographical features and environment in the southern polar region. This paper introduces the strategy of the SELENE 2. Planning forfuture lunarlunar exploration follow-on mission and exploration. It presents the main technical issues faced by designers of lunar exploration robots, a technical of their In Japan, the “Basic Plan for Space roadmap Policy” was estabdevelopment, and the results of studies of a robot lished by Strategic Headquarters for Space Policysystem in 2009. configuration. The report states: “The Government will conduct an exam•

ination of the Japanese original, extensive and long-term the world in Insolving Japan,questions the “Basic Plan for Space and Policy” was of concerning the origin evolution established by Strategic Headquarters for Space Policy in the moon and to investigate the possibility of scientific 2009. The report states: “The Government will conduct exploitation of the moon and utilization of its resources, an examination of the Japanese original, extensive and from the moon perspective of manned on the moonthe that long-term exploration thatactivities is intended to lead will enable in situ informedconcerning judgment.” the To origin plan a and lunar world in solving questions exploration program, “The Panel on Lunar Exploration” was organized in July 2009. The purpose of this panel (Committee) is to study the following issues:

Nine meetings of this panel were held by the end of July 2010. The final conclusions of the panel were reported to the minister of space development in August 2010. The final report of the committee states that, in preparation for a full-range robotic exploration in 2020, a lander and a robot should be sent to the lunar surface in around

2.moon Planning for Lunar exploration that is Exploration intended to lead

Figure 1: Artist concept of a lunar base in south pole Figure Artist concept with an 1. intelligent robotof a lunar base in south pole with an intelligent robot.

• Lunar exploration by 2020: – the objectives of robotic exploration; – research roadmaps; – technical issues; – the spin-off effect on the business market. • Lunar exploration in the long term: – the objectives of robotic and human exploration; – issues to be faced; – international cooperation. Nine meetings of this panel were held by the end of July 2010. The final conclusions of the panel were reported to the minister of space development in August 2010. The final report of the committee states that, in preparation for a full-range robotic exploration in 2020, a lander and a robot should be sent to the lunar surface in around 2015 [4]. This corresponds to the SELENE follow-on mission, which includes a highly accurate unmanned automatic soft landing on the lunar surface. A rover-type robot will then be used to select a suitable site and install observation instruments, including a seismometer. The report also says that we should aim to achieve in 2020 the following outcomes that will lead the world: • lunar base construction and robotic exploration constituting a total travel distance of more than 100 km in several months; • Long-term energy supply by the world’s first lunar renewable energy system; • Japan’s first round-trip to a celestial body that has gravity (bringing back samples). An artist’s concept of the lunar base in 2020 is shown in Fig. 1 and a road map for lunar exploration is shown in Fig. 2.

3.

Environment of lunar polar zones

The moon’s surface is an adverse environment for machines, and designing mechanisms that can operate under such conditions is technically challenging. In the polar regions of the moon, high areas such as crater rims may be permanently sunlit while the bottoms of craters may be permanently shaded, especially in the southern polar region. It is assumed that a lunar base would be sited at a perpetually sunlit location. However, since these locations are in mountainous areas, there are many surface undulations and the surface inclination is expected to range from level to up to 30° (the rest angle of regolith), with an average slope of about 15°. The inner walls of craters are even steeper. Moreover, since the incident sunlight shines almost horizontally, slight surface undulations produce large shadowed domains. For these reasons, the south polar region contains the most areas where the surface receives minimum sunlight. Therefore, it is expected that, even if a low vehicle such as a rover chooses and traverses a path, it cannot receive full sunlight. Moreover, there are wide local temperature variations, ranging between −30° and −50°C in sunlit areas down to −230°C in shadow regions.

4. Missions and means for their realization The major objectives of the deployment of robotics in the SELENE follow-on lunar exploration missions have been set as: • survey for the purpose of lunar exploitation; • scientific exploration;


inner walls of craters are even steeper. Moreover, since the incident sunlight shines almost horizontally, slight surface undulations produce large shadowed domains. For these reasons, the south polar region contains the most areas where the surface receives minimum sunlight. Therefore, it is expected that, even if a low vehicle such

Other options are being studied, and their inclusion will be determined after the international exploration strategy has been clarified. Lunarof surface usingmentioned mobile robots above, a To realize each the exploration objectives number of task elements can be identified, as shown in

Figure 2: A road map for human lunar exploration (JAXA’s vision; not authorized)

Figure 2.

A road map for human lunar exploration (JAXA’s vision; not authorized).

• demonstration of robot technology in outpost construction and operation; • international collaboration. Astronomical observation from the moon is also being considered as a scientific objective. The investigation of soil and bedrock characteristics will be important for outpost construction. Validation of rover technology and trials of position measurement technologies to detect surface movement are also possible mission objectives. These will appear in the late 2010s, in view of Japan’s foreseen participation in human lunar activity. SELENE-X may perform demonstrations of any of the following: • technology for use in outposts, such as the autonomous robot and fuel-cell technology; • the logistics capability to build common landers for both transportation and JAXA’s own robotic missions; • highly sophisticated return of samples of the lunar surface soil to the Earth, including the development of a high-speed reentry capsule. Other options are being studied, and their inclusion will be determined after the international exploration strategy has been clarified. To realize each of the objectives mentioned above, a number of task elements can be identified, as shown in Table 1.

Table 1.

Task elements of lunar rover.

Elements

Objects

Traversal Mapping Measurement Abrasion Observation Pick-up Coring Installation

the target terrain environment and position rocks rocks and soils rocks and soils rocks and soils mission equipments

The following methods can be considered for the analysis of the soil and rock samples collected by a rover. Functions required for the SELENE follow-on robot and future lunar robots are listed in Table 2.

5. Rover technical issues and system configuration The SELENE-X rover is a medium-sized rover whose mass is about 300 kg. The main characteristics of the SELENEX rover for lunar south pole exploration in the 2020s are shown in Table 3. The technical issues that face designers of robots intended to operate on the lunar surface, and their possible solutions are described next.


Battery’s mass

100 kg

Traction

4WD/4WS + Rocker

S.-I. Nishida, S. Wakabayashi Max. velocity

Robot arms

1.4 m/s 2 x 7 DOF

heliostat would have the merits of low mass and high efficiency. In addition, at prospective outposts, it is thought that a power generation tower that generates large amounts of electric power from sunlight and a regenerative type of fuel cell will be essential.

Table 2. Functions required for future lunar robot. Table 2: Functions required for future lunar robot

Table 3.

Characteristics of SELENE-X rover.

Items Mass Max. payload Size Max. power Battery’s mass Traction Max. velocity Robot arms

5.1.

Performance 300 kg 200 kg 2.5 m × 2 m 1 kW 100 kg 4WD/4WS + Rocker 1.4 m/s 2 × 7 DOF

Electrical power and thermal control

As explained above, it cannot be expected that a rover will be continuously exposed to sunlight. The temperatures are always low in the south polar region. This presents a serious challenge in terms of the power supply and thermal control of a rover operating in shadowed regions. The incident sunlight shines almost horizontally, but a solar array raised vertically above the rover may reach above the shadow and be used for photovoltaic power generation. However, it is thought that it would be necessary to raise such a solar array fairly high.

formation, power may be supplied from a lander by a cable, and the rover would be equipped with a reel holding the cable. This appears to be a promising solution, especially for the initial check-out period. However, if power from the lander is used to charge a battery in the rover which then disconnects and moves away to explore, and the battery then runs low, there is a high possibility that, if unexpected difficulties are encountered, a system which must return to the lander to recharge will discharge before reaching the lander, and recovery may become impossible. This solution is therefore not acceptable. However, if the upper part of the lander is a few dozen meters above the moon’s surface and can receive sunlight for long periods, a system which directs the sunlight toward the rover using a heliostat would have the merits of low mass and high efficiency. In addition, at prospective outposts, it is thought that a power generation tower that generates large amounts of electric power from sunlight and a regenerative type of fuel cell will be essential.

5.2.

Communication

Since at present only inadequate detailed geographical feature data exist, such as the undulations of the moon’s surface, it cannot be guaranteed that any generated photovoltaic power will be adequate, and therefore alternative means of power supply must be examined.

The UHF band is suitable for communication between a rover and a lander. The S-band or X-band is suitable for direct communication from earth to a rover while traversing a location that is visible from the earth. It is possible to control a rover remotely from the earth using this communication technology.

If the use of radioisotopes is ruled out, then in areas for which there is no detailed topographical and sunlight in-

When traversing a location distant from a lander, and on the far side of the moon, communication with a rover is


Lunar surface exploration using mobile robots

Table 4.

Sensors for course control system.

Sensors Fiber optical gyro Star tracker Stereo camera Wide field camera Laser range finder

Figure 3.

Allocation of actuators of Rocker-Crawler.

Since regolith has sharp edges and contains many fine particles, it acts as an abrasive and therefore measures should be taken to protect the components that are in contact with the regolith from severe abrasion.

5.4. performed via a data relay satellite that is in lunar orbit or at the Lagrange point.

5.3.

Traction mechanism

Since the thickness of the regolith layer is a few meters or more, the support force of the ground is weak. For this reason, when a rover traverses the moon’s surface, slips and subductions may occur. The regolith is homogeneous and covers the entire surface of the moon, which means that there is little variation in surface characteristics from place to place. Since it is not subjected to the weathering actions of wind or water, its particles are generally angular and sharp. For this reason, its bulk density is low, an average of 1.45–1.55 g/cm3 at a depth of 15 cm according to past research. The force that can be supported by a regolithcovered surface is low compared with terrestrial sands, and the recommended ground support force to be assumed for a run over the surface is set at 1.4 kPa [5]. Because of the low ground support force, it is very important to reduce the ground pressure of the locomotion mechanism in order to avoid getting stuck, which could be fatal for an unmanned robot. Since the moon’s surface is blanketed by regolith, there are a few small-scale bumps and most of the undulations are gradients caused by underlying geographical features. For this reason, the performance over regolith-covered slopes and the traveling ability over rough terrain are thought to be important criteria, and system selection and development work are at an advanced stage. In tests of hill-climbing performance on a slope covered with the regolith, the rocker-crawler type traction system achieved obtained good results. This is based on the low grounding pressure and bundle hardening effect of the crawler belt. Optimization of the belt and suspension mechanisms is at an advanced stage [6] (an allocation of actuators: Fig. 3).

Position determination

Determination of the rover’s own position using signals from a GPS satellite orbiting Earth is difficult on the moon, and since the sun may not be visible over much of a path over the moon’s surface, direction measurements cannot depend on a sun sensor. Dead reckoning based on the odometric information from the traction system would not be sufficiently accurate due to errors caused by the robot sliding over the regolith. Using a radio emitter or optical reference mounted on the lander or a star tracker and a sensor mounted on the rover are effective means of position determination, but rely on an optical or radio line of sight being available. Therefore, determining changes in position by reference to geographical features, using for example range-finding and/or stereo image sensors, is considered essential. Equipping the rover with a stereo imaging sensor is seen as essential, and, since many domains are in shadow, the operation of the sensor will require LED or flash lighting. Obstacles can also be directly detected using a rangefinder, and equipping the rover with a small laser rangefinder to enable three-dimensional measurements is desirable. To address these considerations, the rover course measurement/control system will ideally have the components shown in Table 4, although resource restrictions may force some items to be omitted. A concept of the 2020’s dual arm rover is shown in Fig. 4.

5.5. Operation and control To allow earth-based controllers to monitor the rover, it will be necessary to transmit images at a fixed rate from a forward-facing sensor mounted on the rover. It will also be necessary to construct a continuously updated threedimensional map of the lunar surface on the ground based on images or measurements received from the lander. If, based on this information, operators can transmit detailed


rangefinder to enable three-dimensional measurements he moon's surface is ew small-scale bumps using for isexample range-finding and/orisstereo image sensors, considered essential. desirable. wgradients small-scale bumps caused by sensors, is considered essential. Equipping the rover with a stereo imaging sensor is To address these considerations, the rover course gradients For this caused reason, by the Equipping the rover with a stereo imaging sensor is seen as essential, and, since many domains measurement/control system will ideally have the are in shadow, For reason, ed this slopes and the theNishida,seen S.-I. S. Wakabayashi as essential, and, since many domains are in shadow, the operation of the sensor will require LED or flash components shown in Table 4, although resource din are slopes and tothe thought be the operation of the sensor will require LED or flash a lighting. Obstacles can also be directly detected using may force some items to be omitted. A ntion are and thought to restrictions be development can also be detected using a rangefinder, thedirectly rover with laser concept oflighting. the 2020’sObstacles dualand armequipping rover is shown in Figure 4. a small on and development rangefinder, and equipping the rover with a small laser rangefinder to enable three-dimensional measurements is ormance on a slope Table 4: Sensors for course control system mance on a slope rocker-crawler desirable. type Sensors To considerations, the rover course rocker-crawler typeis address these good results. This Fiber optical measurement/control systemgyro will ideally have the good results. is pressure and This bundle Star tracker components shown in Table 4, although resource essure and bundle t. Optimization ofrestrictions the mayStereo force camera some items to be omitted. A Figure 6: Output of course generation function Optimization the at an advancedofstage. Widedual fieldarm camera concept of the 2020’s rover is shown in Figure 4. at urean3)advanced stage. Laser range finder e 3)contains many fineTable 4: Sensors for course control system nd ddcontains fine thereforemany measures Sensors 5.5 Operation and Control therefore that measures mponents are in To allow earth-based controllers Fiber optical gyro to monitor the rover, mponents that are itinwill be necessary e abrasion. Star tracker to transmit images at a fixed rate from a.Stereo Artist's conception (a)camera Artist’s conception abrasion. a forward-facing sensor mounted on the rover. It will also Figure 6: Output of course generation function a. Artist's conception Wide field camera be necessary to construct a continuously updated

Figure 6. Output of course generation function. Laser range finder three-dimensional map of the lunar surface on the ground based on images or measurements received from the lander. If, based this information, operators can 5.5 Operation andonControl transmit detailed path instructions, themonitor control the of arover, rover To allow earth-based controllers to An outline of the functional assignment of the rover’s remote require atohigh degree of autonomy. However, itwill willnot be necessary transmit images at a fixed rate from control system between the terrestrial and lunar-based a time lag of mounted about eight seconds agiven forward-facing sensor on the rover. Itbetween will alsoa elements is shown in Fig. 5, and the display output of a command signaltobeing sent from earth and a feedback be necessary construct a the continuously updated prototype ground-based course generation function system signal being received rover, considering a three three-dimensional map from of thethe lunar surface on the ground is shown in Fig. 6. secondonsignal round-trip time between the earth andthe the based images or measurements received from moon and addingon signal processing and computing lander. If, based this information, operators can delays, the rover path will need to steer the itself autonomously to transmit detailed instructions, control of a rover 5.6. Manipulation b. Prototype oftotrack module follow the directed path and handle slipping or sliding. will not require a high degree of module autonomy. However, Figure 4: 2020’s lunar rover and a prototype of its track b. Prototype of track (b) Prototype of track module An aoutline of functional assignment ofbetween the rover’s given time lagtheof about eight seconds a of its Although module Figure 4: 2020’s lunar rover and a prototype track the moon’s gravity is only one sixth that of earth’s, remote control system between the terrestrial and command signal being sent from the earth and a feedback a robot arm needs to support the weight of its own wrist or module Figure 4. 2020’s lunarfrom rover prototype of its track lunar-based elements is and shown in considering Figure 5, module. and the signal being received the arover, a three a grasping payload, and to operate. Actuator loads will be displaysignal output of a prototype ground-based course second round-trip time between the earth and the greater than for an armmechanism and drive for mechanism generation function system is shown in Figure 6. Figure 7: Deployment a rover operating in moon and adding signal processing and computing earth’s orbit. Depending on the load, the contact pressure delays, the rover will need to steer itself autonomously to on the bearings is increased by the contact pressure of the follow the directed path and to handle slipping or sliding. reduction gear mechanism and by gravity. Furthermore, An outline of the functional assignment of the rover’s remote control system between the terrestrial and the operating temperature range is very wide and proteclunar-based elements is shown in Figure 5, and the tion is required to prevent regolith dust from entering the display output of a prototype ground-based course mechanism. generation function system is shown in Figure 6. Figure 7: Deployment mechanism for a rover Candidates for the manipulator mechanism include DC brushless motors, which are often implemented in terrestrial robots, mechanisms for other space flight projects (ETS-VII [7], MFD, etc.), and harmonic drives. Since the loads are Figure 8: Testing hill climbinginstalled performance heavy compared to the apparatus on satellites, a direct drive mechanism is not considered suitable. 5.6 Manipulation Figure 5: Schematic diagram of a lunar robot remote ForAlthough structuralthe assembly wiringis(cable moon’sorgravity only construction one sixth thatand of Figure system 5. Schematic diagram of a lunar robot remote control system. control combination of connector) work, positioning to earth’s, a robot arm needs to support the weightrelative of its own wristobjects or a grasping payload, operate. loads other and force controland willtobe neededActuator [8–10]. Since willwork be will greater than for an aarm the be carried out by roverand thatdrive is notmechanism stabilized operating orbit. Depending on the load, the by a scaffold, aearth's grip interface with performance a high grasping stability Figure 8:inTesting hill climbing path instructions, the control of a rover will not require a contact pressure on the[11]bearings is increased by the and positioning marking is essential. high degree of autonomy. However, given a time lag of contact pressure of the reduction gear mechanism and by 5.6 Manipulation about eight between a command signal sent Figure 5: seconds Schematic diagram of a lunar robotbeing remote gravity. Furthermore, operating temperature Although the moon’sthe gravity is only one sixth range that ofis control system from the earth and a feedback signal being received from very wide andarm protection required prevent regolith 5.7. Carriage and earth’s, a robot needsdeployment toissupport thetoweight of its own the rover, considering a three second signal round-trip dust or from enteringpayload, the mechanism. wrist a grasping and to operate. Actuator loads time between the earth and the moon and adding signal To ensure that athan roverfor operating regolith-covered will be greater an arm on andthedrive mechanism operating in earth's orbit. Depending on capability the load, the processing and computing delays, the rover will need to lunar surface has sufficient hill-climbing and contact on the bearings is the increased the steer itself autonomously to follow the directed path and mobility,pressure the ground contact area of rover’s by traction contact the and reduction gear mechanism and For by to handle slipping or sliding. system pressure must be of large the ground pressure low. gravity. Furthermore, the operating temperature range is very wide and protection is required to prevent regolith dust from entering the mechanism. Unauthenticated Download Date | 9/24/15 11:36 PM

161

An outline of the functional assignment of the rover’s remote control system between the terrestrial and lunar-based elements is shown in Figure 5, and the display outputFigure of a 6:prototype ground-based course Output of course generation function generation function system is shown in Figure 6.

onitor the rover, a fixed rate from over. It will also uously updated ce on the ground eived from the operators can ontrol of a rover nomy. However, onds between a h and a feedback Figure 5: Schematic diagram of a lunar robot remote nsidering a three control system he earth and the and computing autonomously to pping or sliding. nt of the rover’s terrestrial and ure 5, and the d-based course igure 6. Figure 7: Deployment mechanism for a rover Figure 7.

Deployment mechanism for a rover.

Lunar surface exploration using mobile robots

Figure 7: Deployment mechanism for a rover

Figure 8: Testing hill climbing performance Figure 8.

Testing hill climbing performance.

5.6 Manipulation

Although the moon’s gravity is only one sixth that of earth’s, a robot arm needs to support the weight of its own a test slope covered with regolith simulant that can be wrist or a grasping payload, and to operate. Actuator loads set at arbitrary incline angles, as shown in Fig. 8. It is will be greater than for an arm and drive mechanism clear from results usingorbit. this equipment that on the the rate of slip-the operating in earth's Depending load, page while traveling over the surface changes according contact pressure on the bearings is increased bytothe the angle of a slope. An reduction undulatinggear geographical feature contact pressure of the mechanism and by model that simulates the surfacetemperature was incorporated gravity. Furthermore, thelunar operating range is into the computer simulation, and a simulation in which very wide and protection is required to prevent regolith the surface friction the coefficient changes according to the dust from entering mechanism. degree of incline was developed. The form of a model and an example of an analysis of a twelve-wheeled crawler traction system with a rocker mechanism are shown in Fig. 9. A dynamical simulation needs to consider especially the flexibility allocation of the suspension of the traction system and servo system parameters in detail.

this reason, as for the traction system, it is desirable that the system be as large as possible within mass constraints. In order to transport such a rover in a lander and to deploy it, it will be necessary to transport it in a somewhat folded state, to unfold it after a landing on the moon, and to 6.2. Measurement / control system testing release it from the lander onto the moon’s surface. For this reason, equipment to lower a rover from a lander and To test the function and performance of measurement sena mechanism itself are needed. An Figureto8:deploy Testingthehillrover climbing performance sors or integrated functions, it is first necessary to simulate example of a rover deployment mechanism is shown in adequately the environment, such as lighting and geoFig. 7. 5.6 Manipulation graphical features, and to qualify the simulation. The test ar robot remote If the rover must return the lander for storage or that of Although the samples moon’s to gravity is only one sixth environment should simulate the incident solar lighting at analysis, a mechanism for needs transferring samples needed earth’s, a robot arm to support the is weight of its own the latitude of the assumed landing site and the interplay on the roveroror lander. If payload, the roverand is equipped with only loads wrist a grasping to operate. Actuator of the light on geographical features. will arm, be greater thanperformance for an arm andofdrive mechanism one robot the required range the robot such as geographical feature map generation operatinglarge, in earth's Depending on the toload, Functions the arm becomes and it orbit. may therefore be difficult and course designation that are assigned to the ground contact pressure on the bearings is increased by the design one arm to meet the requirements of both sample system are included in the operation facility and verified contact pressure of the reduction gear mechanism and by collection and sample transfer. Therefore, these systems by dynamic closed-loop testing in combination with the gravity. Furthermore, the operating temperature range is must be considered individually. dynamical simulator of the payload system. very wide and protection is required to prevent regolith dust from entering the mechanism. The dynamics simulator for dynamic closed-loop testing is already built into the operation facility, and is used as an operation simulator for prior verification of operation commands and procedures. Although various irregular 6.1. Dynamics simulation test fields have been constructed in an indoor laboratory In computer simulations of the dynamics of the rover’s moand are being used for closed-loop testing of measuretion, a technical issue is modeling slippage of the traction ment/control functions, outdoor field testing is also effecsystem over the regolith surface. Quantitative evaluation of tive for verifying measurement/control performance during slope climbing performance was achieved efficiently using long-distance runs over natural features.

6.

Rover testing


a technical issue simulator is modeling slippage of the hion, a high dynamical of the payload system. [11] isis The dynamics for dynamic dynamic closed-loop closed-loop [11] system Thethe dynamics simulator tion over regolithsimulator surface.for Quantitative testing is already built into the operation facility, andisis testing is already built into thewas operation facility, and luation of slope climbing performance achieved b. Simulated rover motion used as an operation simulator for prior verification used as an slope operation simulator for prior verification ofof S.-I. Nishida, S. Wakabayashi ciently using a test covered with regolith operation and procedures. procedures. Although Although various various (Coordinate of center of the body) -covered operation commands commands and h-covered ulantand that can be set at arbitrary incline angles, as Figure 9: Dynamics simulation model of a lunar rover ility bility and wn in Figure 8. It is clear from results using this and a simulation result traction s traction low. For ipment that the rate of slippage while traveling over low. For able that surface rable thatchanges according to the angle of a slope. An onstraints. onstraints. ulating geographical feature model that simulates the otodeploy deploy ar surface was incorporated into the computer at hatfolded folded ulation, and a simulation in which the surface friction n,n, and and to to fficient ace. face. For For changes according to the degree of incline developed. nder and ander and ded.form An The eded. An of a model and an example of an analysis of hown inin shown welve-wheeled crawler traction system with a rocker

chanism are shown in Figure 9. A dynamical nder for ander ulationfor needs to consider especially the flexibility gsamples samples Simulated view (a) Simulated view a.a. Simulated cation equippedof the suspension of the traction system and equipped vo nce range parameters in detail. ce system range

erefore be Measurement efore be ements of ments of Therefore, herefore,

he rover's e rover's ge the e ofof the uantitative antitative achieved achieved regolith regolith angles, as ngles, as using this sing this ling over ing over slope. An lope. An ulates the lates the computer computer ce friction e friction of incline f incline

nalysis of alysis of a rocker a rocker dynamical ynamical flexibility exibility stem and tem and

g g

/ Control System Testing

Figure 10: Prototype of rover in field testing Figure 10.

Prototype of rover in field testing.

References b. Simulated Simulated rover rover motion b. (b) Simulated rovermotion motion (coordinate of center of the body) (Coordinate of of center center of (Coordinate of the the body) body)

Figure 9: 9: Dynamics Dynamics simulation model of aa lunar rover Figure simulation model lunar rover Figure 9. Dynamicsresult simulation model of a lunar of rover and a simulaand a simulation tion result. and a simulation result

[1] [2] [3] [4]

[5] A 1/6 mass prototype of rover in field testing is shown in Fig. 10.

[6]

7.

[7]

Conclusion

This paper described a plan for lunar exploration and the results of work in progress on the technical issues related to future surface exploration by robotic rovers. The south polar region of the moon is currently assumed to be a major candidate for future activities. In order to realize a rover, it Figure 10:toPrototype of rover roverasinafield is important consider the total testing system, taking Figure 10: Prototype of rover in field testingdivision electrical power, communications, night survival, of work, and deployment into account. Work on defining the composition of the total lander and rover system will now progress. Meanwhile, technical development and prototype production of elements such as the rover traction mechanism, geographical feature detection, autonomous control, and measurement sensors will continue in parallel.

[8] [9]

[10] [11]

[12]

SELENE and the lunar farside, Science, 2009, Vol. 323 http://www.selene.jaxa.jp/en/index.html http://www.jaxa.jp/about/2025/index_e.html http://www.kantei.go.jp/jp/singi/utyuu/ tukitansa/100730houkokusho.pdf (written in Japanese) Heiken G., Vaniman D., French B., The lunar sourcebook, Cambridge University Press, Cambridge, 1991 Nishida S., Wakabayashi S., A study on mobility and control of the SELENE-2 Lunar Rover, IAC2008, Glasgow, 2008 Oda M., Nishida S. et al., Onboard local compensation on ETS-VII space robot teleoperation, AIM’99, Atlanta, 1999 Nishida S., Yoshikawa T., Reflector assembling in orbit by a joint torque controlled space robot, ISR2004, 2004 Nishida S., Yoshikawa T., Hirabayashi H., A new space robot end-effector for on-orbit reflector assembly, DIS2005, NGI-01, Marseiu, 2005 Nishida S., Yoshikawa T., On-orbit assembly of large reflector, ICARCV2006, Singapore, 2006 Kamimura H., Nishida S., Three dimensional color marker for on-orbit assembling using space robot, IAC2004, Vancouver, 2004 Shiraki K., Kubo T., Nishida S., Simulation for developing JEM remote manipulator, Proc. of i-SAIRAS’92, 1992
