Development of aPersonal Service Robot with User-Friendly Interfaces 433 refrigerator A B C D E X Y Fig.8. Anominal trajectory. side view door top view door Fig.9. Tolerance for segmentCDinFig. 8. We,therefore, develop anovel teaching methodfor amobile manipulator which exists in between the above twoapproaches. In the method,the user teaches the robotanominal trajectory of the hand and its tolerance to achieve the currenttarget task. The givennominaltrajectorymay be infeasible due to the structural limitation of the robo t; so the robo ts earches for af easible one within the tolerance. Only when the robot fails to find anyfeasible trajectory,itplans amovement of the mobile base, by using the redun danc yp ro vi ded by the mobile base as another toleranc e. The teaching metho di sw ell intuiti ve and does not requir em uch user’ se ff or tb ecause the use doesnot have to consider the structurallimitation of the robot in teaching.At the same time, the metho dd oe sn ot assume ah igh recogn ition and inferen ce ability of the robot because the givennominal trajectoryhas much information for motion planning; the robot does not need to generate afeasible trajectoryfrom scratch. The follo win gs ubsections ex pl ain the teaching metho d, using the task of openin g the door of arefrigerator as an example. 4.1 Nominal Trajectory Anominaltrajectoryisthe trajectoryofthe hand pose (position and orientation) in a3 Do bjec t-centered coo rdinate system. To simplify the trajectory teaching ,w e currently set alimitation that atrajectoryofhand position is composed of circular and/or straight line segments. Fig. 8shows anominal trajectoryfor opening adoor, composed of straight and circular segments on some horizontal planes; on segment CD, the robot roughly holdsthe door,while on segment DE, the robotpushes it at a different height. The hand orientation is also specified as shown in the figure. 4.2 Tolerance Atolerance indicates acceptable deviationsfrom anominal trajectorytoperform atask; if the hand exists within the tolerance overthe entire trajectory, the task is achie vab le. Au ser teaches at oleranc ew ithout ex plicitly considerin gt he structural limitation of the robot. Givenanominaltrajectory anditstolerance, the robotsearches for afeasible trajectory. 434 J. Miura et al. nominal trajectory generated trajectory infeasible region via point tolerance Fig.10. Example feasible regions. via points trajectory Fig.11. Afeasible region. The user sets atolerance to each straight or circular trajectoryusing acoordinate system attached to each point on the segment. In these coordinate systems, auser can teach atolerance of positionsrelatively intuitively as akind of the width of the nominal trajectory. Fig. 9shows an example of setting atolerance for circular segmentCDinFig. 8, which is for openingthe door. 4.3 Generating Feasible Trajectories The robotfirst tries to generate afeasible trajectory within agiven tolerance. When the robotfails to find afeasible one, it divides the trajectoryinto sub-trajectories such that each sub-trajectory can be performed withoutmovement of the base; it also plans the movement between performing sub-trajectories. On-line Trajectory Generation The robotsets via points on the trajectorywith acertain interval, and generates afeasible trajectorybyiteratively searching for feasible hand poses for the sequenc eo fv ia poin ts. This trajectory genera tion is per - formed on-line because the relative position between the robotand the manipulated objects may vary from time to time. The robot estimates the relative position before trajectory gene ration. The pre vi ously calculated trajectories are used as guid es for efficiently calculating the current trajectory. Fig. 10 illustrates howafeasible trajectoryisgenerated; small circles indicate via points on the gi ve nn om inal trajectory ,t wo dashed lines indicate the boun dary of the tolerance, the hatched regionindicates the outside of the rangeofpossible hand poses. Afeasible trajectory is generated by searchingfor asequence of hand poses which are in the tolerance and near to the givenvia points. In the actual trajectory generation, the robotsearches the six dimensional space of hand pose. Duringexecuting the generated trajectory,itissometimes necessary to estimate the object position. Currently,wemanually give the robotaset of necessary sensing operations for the estimation. Trajectory Division Based on Feasible Regions The division of atrajectory is done as follows. Foreach via point, the robotcalculates aregion on the floor in the object coordinates such that if the mobile base is in the region, there is at least one feasible hand pose. By calculating the intersection of the re gion s, the rob ot determin es the re gio no nt he floor where the robo tc an mak et he hand follo wt he Development of aPersonal Service Robot with User-Friendly Interfaces 435 A B C D E V X Y Fig.12. Example feasible regions. mobile base 6 DOF arm with hand laser range finder hand camera main camera 3-axis force sensor host computer Fig.13. Our service robot. entire trajectory.Such an intersection is called a feasible region of the task (see Fig. 11). The robot continuously updates the feasible region, and if its size becomes less than acertain threshold,the trajectory is divided at the correspondingvia point. Fig. 12 sho ws ex amp le feasible re gio ns of the trajectory of openin gt he doo rs ho wn in Fig. 8. The entire trajectory is di vided into tw op arts at poin t V ;t wo corre spondin g feasible regions are generated. 5Prototype System and Experiments Fig. 13 sho ws our person al service rob ot. The robo ti sa self-con tained mobile manipulator with various sensors. In addition to the above-mentioned functions, the robotneeds an ability to move between auser and arefrigerator.The robot uses the laser range finder (LRF) for detectin go bstacles and estimating the eg o- motion [8]. It uses the LRF and vision for detectingand locating refrigerators and users. Fig. 14 shows snapshots of the operation of fetching acan fromarefrigerator to auser. 6Summary This paper has describedour personal service robot. The robot has user-friendly human-robot interfacesincluding interactive objectrecognition,robust speechrecog- nition, and easy teaching of mobile manipulation. Currently the twosubsystems, object and speech recognition and teaching of mobile manipulation, are implemented separately.Weare nowintegratingthese two subsystems into one prototype system for moreintensive experimental evaluation. Acknowledgment This research is supported in part by Grant-in-Aid for Scientific Research from Min- istry of Education, Culture, Sports, Science and Technology,and by the Kayamori Foundation of Informational Science Advancement. 436 J. Miura et al. approach open grasp close carry hand over Fig.14. Fetch acan from arefrigerator. References 1. Morpha project, http://www.morpha.de/. 2. R. Bischoff. Hermes –ahumanoid mobile manipulator for service tasks. In Proc. of FSR-97,pp. 508–515, 1997. 3. R. Bischoffand V. Graefe. Dependable multimodal communication and interaction with robotic assistants. In Proc. of ROMAN-2002,pp. 300–305, 2002. 4. M. Ehrenmannetal. Teaching service robots complextasks: Programming by demon- stration for workshopand household environments. In Proc. of FSR-2001,pp. 397–402, 2001. 5. K. Ikeuchiand T. Suehiro. Toward an assembly plan from observation part i: Task recognition with polyhed ral objects. IEEE Tr ans. on Robotics and Au tomat.,V ol. 10, No. 3, pp. 368–385, 1994. 6. Y. Makihara et al. Object recognition supported by user interaction for service robots. In Pr oc. of ICPR-2002,p p. 561–5 64, 2002. 7. Y. Makihara et al. Object recognition in various lighting conditions. In Proc. of SCIA- 2003,pp. 899-906, 2003. 8. J. Miura, Y. Ne gishi, and Y. Shirai. Mobile robot map generation by inte grating omnidi- rectional stereo and laser range finder.In Proc. of IROS-2002,pp. 250–255, 2002. 9. J. Pineau et al. Towards robotic assistants in nursing homes: Challenges and results. Robotics and Autonomous Systems,Vol. 42, No. 3-4, pp. 271–281, 2003. 10. N. Royetal. Towards personal service robots for the elderly.In Proc. of WIRE-2000, 2000. 11. M. Takizawa et al. Aservice robot with interactive vision –object recognition using dialog with user –. In Proc. of Workshop on Language Understanding and Agents for Real Wo rld Inter action,p p. 16-23, 2003 . AnEnhanced RoboticLibrary System for anOff-Site Shelving Facility Jackrit Sut hako rn 1 ,S ang yoo nLee 2 ,Y u Zhou 2 ,S ayeed Cho udh ury 3 , and Gregory S.Chirikjian 2 1 Department of MechanicalEngineering, Mahidol University,Bangkok,Thailand song@jhu.eduor jackrit@trs.ac.th 2 Department of MechanicalEngineering The Johns Hopkins University,Baltimore,Maryland 21218 USA slee@konkuk.ac.kr, yuzhou@titan.me.jhu.edu,gregc@jhu.edu 3 DigitalKnowledgeCenter of the SheridanLibraries The Johns Hopkins University,Baltimore,Maryland 21218 USA sayeed@jhu.edu Abstract. This paper describes our continued workofa unique robotics project, ComprehensiveAccess toPrinted Materials (CAPM), within the context of libraries.As libraries provide agrowing arrayof digitallibrary services and resources, they continue to acquirelarge quantities of printed material. This combined pressureofproviding electronic and print-based resources and services has led to severe spaceconstraints for many libraries, especiallyacademic researchlibraries.Consequently,manylibraries havebuilt or plan tobuild off-site shelving facilities toaccommodateprinted materials.Anautonomous mobile robotic library system has been developed to retrieveitems from bookshelves and carry them to scanning stations located in the off-site shelving facility.This paper reviews the overall design of the robot system and control systems,and reports the new improvement in the accuracy of the robot performance; in particular, the pick-upprocess. 1Introduction As libraries provideagrowingarray of digitallibrary services and resources, they continue toacquirelarge quantities of printed material. This combined pressureof provid ing electron ica nd prin t-ba sed resour ces and services has led to severe s pace constraints for many libraries,especially academic researchlibraries.Consequently, many libraries havebuilt orplan tobuild off-site shelving facilities toaccommodate printed materials.However,given that theselocations arenotusually withinwalking distanceof the main library,access to thesematerials, specifically the ability to browse,is greatly reduced.Libraries with suchfacilities offer extensivephysical delivery options from thesefacilities, sometimes offering multiple deliveries per day. Even with suchdelivery options, the ability tobrowsein real-time remains absent. The goalof the CAPMProject is tobuild a robotic, on-demand and batch scanning system that will allow for real-timebrowsingofprinted materials through a web interface. Weenvisage the system will workas follows:anend user will identify that amonograph is located in anoff-sitefacility.The user will engage the CAPM system that,in turn, will initiatea robot that will r etrieve the req uest ed item .The robot will de liver thi s item toano the r robotic s ystem that will op en the item and t ur n t he page s automatically.By usingexisting scanners,opticalcharacter recognition (OCR) software,and indexing softwaredeveloped by the DigitalKnowledgeCenter (a researchand development unit of the SheridanLibraries at Johns Hopkins), the S. Yuta et al. (Eds.): Field and Service Robotics, STAR 24, pp. 437–446, 2006. © Springer-Verlag Berlin Heidelberg 2006 438J.Suthakorn et al. CAPM system will not only allow for browsingofimages of text,but alsofor searchi ng and analyz in go ff ull- text ge nerated from t he im ages. The details of the me chanical structur e, t he navig ation sys tem ,contr ol and software, simulations,experiments and results of the robot werepreviously described in [1].This pape r focus es on im provem ent in the accuracy of en tiredelivery proced ureof the robot system ,in particular, t he pi ck- upo fboo kca ses, w hi le t he future work, will beconcentrated on developing the robotic system tocomplete the rem ain in gprocesses. Since the CAPM robot is designed to workinanoff-site shelving facility that belong to the Johns Hopkins University, severalassumptions in the design aremade based on the actualenvironment of this facility.All the paths in the facility are assumed tobe smoothand flat.Eachbookis assumed tobe stored in aspecialcase, whichhasapair of wing-likehandles for engaging withapassivegripper.Itis assumed that eachitem is stored in a specifically designedcaseand arranged side-by- sidewitha small in-between gap. Finally,abarcode is attached toeachcase. Currently at the MoraviaPark shelvingfacility,after receiving a request,alibrary officer at the facility willdriveaportable personnel lift to retrieve the requested item toits location,and thenbringit toa waitingareafor the next scheduled transportation Then,abatchof requested items is transferred to the main library.In the same manner, our robot will beinitially parkedat the docking station until anitem is requested.The robot is equip ped withadatabase system of booklocations and aglobalmapof the off-site shelvingfacility.After receiving a request, the robot will autonomously run along aknownpath to the booklocation and retrieve the requested item from the shelf. Then, the robot willcarry the item back to the scanning stationand thenreturn to the docki ng statio n. This is not the first timea robot has been built toperforma specific service function.In1995,Hansson introduced anindustrial robot in aSwedishlibrary [2]. Safaric[3]presented anexample of a telerobotcontrolled viaInternet [4].Byrd introduced a successful service robot used to survey and inspect drums containing low-level radioactive waste stored in warehouses at Department of Energy facilities [5]. The CA PM sys tem di ffers from other exist in g sys tem s in t he fo llo win g w ays . First, the system retrieves in di vid uali tem s,a sop po sed t oboxes of item s, s uchas the system at the CaliforniaStateUniversity atNorthridge [6].Second, the CAPM system does not assume anexisting or fixed shelving and spacearrangement.This flexibility will allow it toworkinmany diverseenvironments.Third, the CAPM retrieval robot is anautonomous system.Fourth, the economicanalysis by a collaborating researchgroupin the Department of Economics at Johns Hopkins University has verified that a relatively inexpensive robotic system is cost-effective, especially in comparison topotentialbenefits.Finally, the page-turning system, tobe built in the future, will accommodatea wide variety of paper types and materials. In subsequent sections of thispaper, we report the design,control systems, experiments and results of anautonomous roboticlibrary system for anoff-site shelvingfacility.Sections 2and 3 briefly review the robot design, the robotcontrol sy st em s and s of tw areand navig atio n s ys tem (detaile dd escrip tio ns ca nbef ound in [1, 6].) Section 4explains improvement in the accuracy of entiredelivery procedureof the robot system.We thenreport experiments and results in Section 5. AnEnhanced RoboticLibrary System foranOff-SiteShelving Facility 439 2 Hardwareo f the RoboticS ystem 2.1 MechanicalStructure This section presents designs and descriptions of twomajor components of the CAPM library robot: the manipulatorarmand the locomotiondevice. 2.1.1 Manipulator arm system Inorder to retrievebooks from bookshelves and carry them to the scanning station,a specificmanipulator arm system was designed.Sinceeachbookshelf is 10-foot-high, a vertical translation system (VTS) was used tomove the robot manipulator to different altitudes.The VTS is a sliding rod withanelectricmotor for driving alead- screwed rod.Anenhanced commercial 6-DOF robot manipulator, the F3 made by CRS Robotics,Inc.,is affixed toaplatformwhichis apart of the vertical translation system (See Figure1.) Webuilt and installed apassivegripper to the end-effector of the robot manipulator.The gripper is used topassively grasp the bookcase. The structures of the gripper and bookcases weredesigned tofittoeachother.Abarcode scanner is installedinside the gripper in order torecognizeand ensure the precision of picking a requested item. 2.1 .2 Locom otion device The locomotiondeviceis responsible for the gross motionof the robot.Wehave mo di fi ed acom me rcial s erv o-con tr ol le dm obile robotpl atform, t he Lab matem ade by Help mateInc.Analumi num- allo y cart is built and attached to the Labmatem obile platform. This cart isused to store the robot manipulator controller and the power source while the Labmatemobile platformis used as the baseof the manipulatorarm system.A ranging sensor system was installed on the mobile platform tocollaborate and improve the navigation system.All electronicdevices used tocontrol the vertical translation system and sensor systems wereinstalled on the mobile platform. Because of the in stallatio nofapo wer sour ceo nboard, t he robot does no t req uireane xt ernal po wer lin e w hile worki ng .Fig ure1 shows the overall mechanical s tr uctu reo f t he library robot. Inour work,due tolimits of sensorperformance,8 sensors are used:4 sonar sensors and 4infrared sensors.One sonar and oneinfrared sensor arepaired together toget eachof the 4 sensor readings needed. This is done becauseof the distance measuring limits whicheachhave. The Polaroid6500 sonar sensorhas a rangeof15- 1067 cm while the SharpGP2D02 infrared sensor has a range of 10-80 cm. It canbe observed thatby using these two sensors combined, wecanachievea range of 10- 1067 cmof reliable distancemeasurement.Each sensor is controlled and interfaced to the main computer viaamicro-controller (BASIC Stamp II). 3Control and Software All the processes and activities of t he syst em arecontr olled by ano nboardIntel PentiumII laptop. The control systems of the library robot consist of several sub- controllers: the control system of the VTS, the control system of the robot 440 J.Suthakorn et al. manipulator, the control system of the mobileplatform, the high-level control system of the library robot,a nd the contr ol software. 3.1 Control of the VTS The VTSis required in the library robot tomove the robot manipulator todifferent altitudes.Alead screw lift sys tem was selected and mo difie d tobeamajor com pone nt of the VTS.Toenhance the lift system so that it could successfully beimple mented as part of anautonomous roboticlibrary system,afeedbackcontrol system was integrated.This feedbackcontrol system functions to:1) determine the altitude of the VTS platform, 2) send the real-time ranging information to the high-level controller, and 3) receiveacommand from the high-level controller tocontrol the motion of the VTS platforminorder to reachadesired altitude. The feedbackcontrol system of the VTS consists of aBASICStamp II micro-controller,aninput/output serial communicationport,a range sensor system,and aVTS directionalcontroller. Fig.1. The RoboticLibrary System 3.2Control of the Robot Manipulator The six-axisF3 robot system manufactured by CRSRobotics Incis used as the control system of the robotmanipulator.Articulated joints provide the F3 arm with six degrees of freedom,and absoluteencoders mounted on the motor shaft in each joint providepositionalfeedback to the controller.The F3robot arm uses the Cartesiancoordinate system. Control programs were written in the C++ language and downloaded to the controller.Control programs use the ActiveRobotinterfacedeveloped by CRS robotics and include twoobject classes of the ActiveRobot interface:one providesthe application with the main interface to the robot system,and the other enablesthe application tocreateand modifyrobot locations.The input variables to the control programs are the speed and the positionand orientationof the finallocationof the end-effector.The contr olle r provid es the com put atio nofi nversek in em atics. Sonar/IR Sensor System for Robot Navigation Locomotion Device Robot Manipulator IR Sensors for Manipulator Motion Planning Gripper Vertical Translation Device AnEnhanced RoboticLibrary System foranOff-SiteShelving Facility 441 3.3 Control of the MobilePlatform The locomotivedevice,or the mobile platformLabmate,has the drive system microprocess or that the u ser' s host com put ercomm unicates with t hrough anRS -23 2 serialport.The host computer always initiates communicationsbetween the host computer and the Labmate. The Labmateuses aCartesiancoordinate system for position con trol .The coo rdi nate s ystem isagl oba l referen ce t hat is in itialized at power upand reset.Odometry (dead reckoning) is the practiceofcalculatingposition from wheel displacements.The Labmatecontrol system depends on encoders mounted on each wheel tokeep trackofposition. Control programs arebasically composed of commands that direct the Labmate toaparticularlocation. Weassume here that the entiremapof the workspaceis stored in the formofalook-up table in the memory of the Labmate. Ifadestination is given, the Labmatecomputes the direction from the current locationbyreferring to the look-up table. Tocompensate for the errors,we used twokinds of sensors: ultrasonicranging sensors and infrared sensors.For more successfulnavigation of anautonomous vehicle over extended distances, references to the external world at regular intervals arenecessary.Figure2 illustrates the diagramof the library robot controls. 3.4 Control Systems of the Library Robot Wecall the main control system of the library robot the “high-level control system”. This control system consists of aPentiumII 233 MHz computernotebookand a serial po rt spl itter hub.The com put er no teboo kfunctio ns as the cen tralp rocessin guni t of the library robot,a nd it com muni cates toe very subsy stem through the s erialports. 3.5 Control Software The control softwareisdesignedbased on the ideaof eventdriven programming. Principally, the main control programcontrols the mobile platform, the vertical translation device,and t he armthrough serialp orts.Whe n t he main con trol prog ram beg in s to r un,it in itializes the serialports of the com put er at first and starts the even t listeners for all the serialports.Then the main control programleaves the control to the listeners.It is actually theseevent listeners that control the movement of the platform, the vertical translation device,and the arm. Basically,eachevent listener will monitor the statusof one serialport.Once the status of that port changes, the listener will judge what kind of event happens and executeacorrespondingfunction. We use the word‘lift’ interchangeably with‘the verticaltranslationdevice’. Animportantpropertyof event driven programming is that the executionorder of the functions is not fixed,it depends on the need toexecute. This property is suitable for the sensor driven system of the library robot.In totalfour event listeners are created. They monitor the status of the platform, the lift, the arm,and the sensors respectively.Figure 3 shows the software structureof the robot. 4Improvement in the Accuracy of the Book Pick-UpProcess Toassist t he mo tio np lanni ng of t he library robot, t he com pl ete w orki ng process of the library robot was sim ulated us ing 3DS MAX.Based on the sim ulatio n,a com pl ete pathwas generated,and experiments wereexecuted to test and adjust the performance of the robot.To simplify the implementation,amap-based scheme was employed to 442 J.Suthakorn et al. control the mobile platform. Afixed globalcoordinate system is defined withits orig in at the do cki ng spot.The po sitio ns of the in terme di ate stop s and the de stin ation wered ef in ed in t he gl obal s ys tem .Ano ptim alp ath w as chosen t ocon ne ct the cur ren t stop and the next stop.The major problemappearing in the experiments is positioning err or.It was found t hat the positio ni ng err or was clo sel y related t o t he mo ving spe ed of the platform. If t he spee d w as low, the mo tor may lose s om e steps becauseo f the certain heavy overall load. If the speed was high, the platformmay deviatefrom the desired pathat the turni ng corne rs beca useo f t he in ertia.After afew adj us tme nts of the speed setting, the positioning was improved considerably. Fig.2. Diagramof the library robot controls. Fig.3. Softwarestructureof the library robot. Toenhance t he accuracy of pi ck-upprocess in the r obot manipulator con trols, w e in teg rated ani nf rared sen sing s ys tem t o t he mani pulator.The in frared sen sing sys tem consists of twoinfrared sensors (SharpGP2D02), the sensor controller,and an input/output serialCOM port tocommunicate with the high-level controller.The two infrared sensors wereattached to the fingertips of the end-effecter,and arecontrolled by acircuit controller made of aBASIC Stamp II microchipand other required [...]... experiment set-up of the robotic manipulator tests, and in the following part the experimental results will be described There are no failed trials (out of 51 trials) in this experiment Figure 8 shows the step-by-step picking up process of the experiment The robot manipulator started at its initial position and begins scanning from far-left to far-right positions After the algorithm determines and finds... through a combination of random motion and fixed motion pattern Fraunhofer Institute AIS, Germany modified commercial platform KURT2 with differential six-wheeled drive system, 454 E Prassler, M H¨ gele, and R Siegwart a 10 ultrasonic range sensors, commercial electric broom, coverage by locomotion on parallel tracks and random motion Short Circuits Robotics Club, Ireland/Czech Rep self-designed platform with... Germany self-designed differentially driven vehicle, high resolution range sensor allows differentiating between carpet, vase and other objects Hertwig Andre (private inventor), Germany self-designed vehicle with differential drive, self-designed brushing and vacuuming unit, reflecting light sensor of collision detection, coverage by combination of random motion and locomotion along a meander 3.2 The... of 33% of the achieved scores 2.8 Handicaps It was in the spirit of this first contest to allow a maximum level of creativity in the design of the robots According to the contest rules self- designed robots with self-designed cleaning technology were admissible as much as off-the-shelve experimental platforms equipped with some off-the-shelve cleaning mechanism High-tech platforms with expensive sensor... Wing-Liked Handle Fig.5 Motion planning of the robotic manipulator in the pick-up process Fig.6 Control architecture of the manipulator’s sensing system 5 Experiment and Results of the New Improving Pick-Up Process Experiments on the robot manipulator were conducted to determine the accuracy of the pick-up process performed by the manipulator and its feedback control system The experimental set-up,... as cleaning and that a cleaning robot contest could attract only roughly as many students as a contest of soccer S Yuta et al (Eds.): Field and Service Robotics, STAR 24, pp 447–456, 2006 © Springer-Verlag Berlin Heidelberg 2006 448 E Prassler, M H¨ gele, and R Siegwart a playing robots Their worries were unjustified The contest became a big success Fifteen teams from 10 countries worldwide participated... results While these points outline the specific benefits and qualities of CAPM, it is important to note a ultimate goal of this project The CAPM Project will introduce robotics into the library and, perhaps more importantly, digital library context As robotics have provided great impact and utility within manufacturing and, increasingly, computer-assisted surgery, it is possible that similar gains will... offthe-shelve components were scored differently, however (see Section “Handicaps”) The following auxiliary devices were allowed: • docking stations and • external movable and removable reference stations and • navigation support systems Non-movable installations were not allowed 2.5 Rules Besides the above requirements the teams and robots had to obey a number of rules to assure a fair contest: • one... 446 J Suthakorn et al Fig.8 Picture shows step-by-step in pick-up experiment: 1) Manipulator starts at the process's initial position, 2) Manipulator starts the scanning process from the far-left position, 3) Manipulator finishes the scanning process at the far-right position, 4) Manipulator begins picking up the bookcase 5) During the pick-up process, and 6) Manipulator successfully picked up the... data EPFL, Autonomous Systems Lab, Switzerland self-design platform using two SICK 2D laser range finder, differential drive mechanism, micro-fibre cleaning towel mounted at the bottom of vehicle, optimal coverage by locomotion on parallel tracks in a partially mapped environment FAW Ulm, Germany self-design robot with differential drive and tactile sensing, self-designed wetcleaning unit, position estimation . design, control systems, simulations,experiments and results werepresented.An implementation usingIR sensors and anew algorithm toenhance the accuracy of an operation process,book pick-upprocess, was described ,and. experiment set upfor testing the roboticmanipulator. 446 J.Suthakorn et al. Fig.8. Picture shows step-by-step in pick-upexperiment:1) Manipulator starts at the process&apos ;s initialposition, 2)Manipulator. mo tion pl ann ing of the robotic manipulatorin the pick-upprocess.The book-position-scanning process is aprocess tocorrect the book-positioning errors.This allows the book tobeplaced in a specific ranging