Industrial Electrical Engineering and Automation - IEA
Contents
1. Figure 6 k The menu for showing the result data The user can select original data interpolated data or mean value data The VBA software consists of a main module and several sub modules see Appendix A The modules control the graphical user interface and handle the communication with the robot code It collects the data from the robot and force torgue system and finally processes that data and presents the data to the user It also handles errors that can result from e g communication errors and numerical inconsistencies In essential it follows the desired measurement procedure in Chapter 2 2 When a test is executed the VBA software transfers the necessary in data to the robot system and executes the robot program that in turn controls the robot The robot arms moves to a predetermined location close to the selected test object These locations have to be taught with the function Teach by the operator before the tests see Chapter 3 2 In case of a push button it starts to push on the button and at the same time samples the force When the force reaches a predetermined maximum value when the button is fully pressed down it terminates the test and transfers the data back to the VBA software It the test object is a rotary switch the robot arm starts to turn the switch and at the same time samples the torgue When the turning angle reaches a predetermined maximum value it terminates the test and transfers the data back to2. When testing rotary switches with distinctive notches it might be better to use a servo gripper since it stays in contact with the switch at all times The fork shaped tool looses contact every time the switch snaps after the notch An alternative could be to use a tool that precisely fits the rotary switch handle There are five standard transducers available from ATI with measurement ranges from 12 N to 5000 N suitable for different measurement situations In this thesis work a more sensitive transducer would have been preferable for better resolution but the Gamma 15 50 transducer was the one at hand 8 2 Specific Challenges It has been a very challenging but interesting work One of the first challenges was to find out the different force and torque variables that were measured by the transducer and the associated coordinate system One subject that accounted for many hours of trial and error was the communication protocol DDE between the VBA software and the robot controller Since the sampling is performed in real time during the robot motion the synchronization of all involved processes was crucial In order to stay synchronized a solution was to check if the robot system was busy or ready for new inputs and accordingly send ready signals to the processes The movement of the buttons or switches during the associated sampling of position and force or torque data could not be done in predefined steps Instead the movement wa
3. Data Diagramrubrik Titel Vinkel Moment Diagram Axlar xlKategori Axelrubrik Titel Vinkel o Axlar xlV rde Axelrubrik Titel Moment N cm End With End With End Select End Sub 54 Sub NyttTestObjekt Knapp Dialogbladlista Dialog3 Visa End Sub Sub NyttTestObjektOK Knapp Skapar nytt testobjekt och f rbereder kalkyl bladslistor GammalS kv g C crs userprog prog_1 excel GammaltFilNamn Kalkylbladlista Data Celler 3 2 Val AktivArbetsbok SparaKopiaSom GammalS kv g GammaltFilNamn XLS Kalkylbladlista Data Celler 3 2 Val Dialogbladlista Dialog3 Redigeringsrutor 17 u Tex t Kalkylbladlista Data Celler 4 2 Val Kalkylbladlista Info Celler Dialogbladlista Dialog3 NedrullbaraListrutor 24 Val 5 3 Val Kalkylbladlista Data Celler 5 2 Val Kalkylbladlista Data Celler 6 2 Val Dialogbladlista Dialog3 Redigeringsrutor 28 u Tex t Now InitTyp Kalkylbladlista Data Celler 4 2 Val Dialogbladlista Dialog3 D61j AktivArbetsbok Bladlista HuvudMeny Aktivera Kalkylbladlista Data Omrade Area 11 2 500 3 RaderaInnehall Kalkylbladlista Data Omrade Area 7 2 8 2 RaderaInnehall Kalkylbladlista Interpolerad Data Omr de Area 11 2 500 3 RaderaInnehall Kalkylbladlista Interpolerad Data Omr de Area 7 2 8 2 RaderaInnehall Kalkylbladlista Accumulerad Data Omr de Area 13 2 500 4 RaderaInnehall Kalkylb
4. GOSUB FON GOSUB ERASER BUSY 0 STOP FON TXT Program Name FON Initializes the Force and Torque sensor SENSOR FORCE INZ initialize the sensor memory SENSOR FORCE ON turn on the sensor sampling RETURN ERASER TXT Program Name ERASER Erases all values in the array DVAR and the array DLOCN N 1 DO 60 DVAR KRAFT IN DLOCN LOCAT N UNTIL N gt MAXNUMB MAXNUMB 1 RETURN MAX TXT Program Name MAX Collects the maximum forces that was measured MAXX FORCE 42 MAXT FORCE 45 TOP AA se EE we RECOVER TXT Program Name Recover Resets the system PASSWORD 255 CLINSPD 40 0 PASSWORD 0 PushButton TXT Program Name PushButton This program does the actual measurement of a push button designated Object The operator have to define a start position PBStart Object and an end position PBEnd Object for each test object BUSY 1 POSE FUN GOSUB FON SENSOR FORCE FRAME RAW MAXFOZ MAXCONOZ SPEED 50 MOVE PBStart Object Move to start position FINISH SENSOR FORCE CALIBRATE X 9 5 MAXFN1 Converts from N to Oz 3 3 96 MAXCONN Converts from N to Oz 6 5 61 STARTF FORCE 0 SPEED 10 MOVE PBEnd Object Start pushing on the switch TROIS AAR KAA IK ARIK IKE IK IK HK KK SAFETY Interrupts if anything in the path if force more than max allowed force c 0 DO IF ABS
5. 15 50 transducer Figure 4 a and Figure 4 b is composed of six compact monolithic silicon strain gage elements The sensor is able to simultaneously measure all of the six components of force and torque The analog signals from the strain gages are low pass filtered to prevent aliasing multiplexed and amplified in the transducer and sent to the F T controller via an electrically shielded cable The transducer is mounted between the tool flange on the end of the last link of the robot arm and the tool The sensor used in these tests is capable of measure forces up to 130 N with a resolution of 0 10 N in the z direction and 65 N with a resolution of 0 05 N in the x and y directions Torque is measured up to 5 Nm with a resolution of 0 003 Nm Furthermore it has an overload protection able to withstand a force slightly more than 1 kN and a torque of 90 Nm 4 2 The F T Controller F T Controller Figure 4 c The primary function of the F T controller is to convert analog strain gage signals from the transducer to digital Cartesian force and torque components The low pass filtered multiplexed and amplified strain gage signals is transmitted from the transducer to the F T controller s 12 bit A D converter 21 The now digitized signal is then sent forward to the CPU which further processes the data and then transfers the data via a 38400 bps RS 232C interface to the robot controller Since the A D converter has 12 bits the final re
6. 3 a is composed of several essential parts The robot arm is described in Section 3 1 The robot controller unit and the operating system that resides inside it and regulates the robot arm are covered in Section 3 2 The user available software is treated in Section 3 3 and finally the host computer that can be utilized as an extended terminal to manage the software and supervise the system is described in Section 3 4 Figure 3 a The robot system 465 from CRS Robotics Corporation 3 1 The Robot Arm The articulated robot arm A465 1 seen to the right in Figure 3 a has six rotating joints which entails to six degrees of freedom DOF Figure 3 b The word articulated means provided with joints 2 The actual definition of DOF reads as follows DOF is one of the variables maximum number of six required to define the motion of a body in space 3 The elementary motions that the body can have in relation to a coordinate system are translation along the coordinate axis and rotation around the axis In the case of a Cartesian coordinate system defining the three dimensional space there are three orthogonal axes and thus a maximum of six degrees of freedom Since the arm has six DOF it means that the end of the arm where the tool is mounted is able to reach any point in the three dimensional space and in any direction as long as the point is inside the working area Figure 3 c of the robot The number of joints however
7. At any given time the controller can be controlled from only one source The most basic method is to use the teach pendant Figure 3 e This is a hand held robot control terminal that provides a means to move the robot edit and run robot programs It has a four line LCD display for presenting and editing programs locations variables and for checking the status of the robot It is also equipped with an alphanumerical keyboard an emergency stop switch and a live man switch if released the live man switch will stop the robot motion The main use of the teach pendant however is to move 16 the robot manually in order to teach the robot locations which will then be referred to in robot programs These locations can be stored together with programs and variables in the controller s battery backed RAM memory 10 with room for 256 kB Figure 3 e The teach pendant The front panel can be utilized to start or stop program execution homing and for emergency stopping The other two methods to manage the robot controller the use of a terminal emulator on a host computer or a robot program that is executed within in the controller will be described in the following sections 3 3 System Software The programs that can be executed in the robot controller are written in RAPL 2 which is an abbreviation for Robot Application Programming Language 11 The language is line structured and similar to simple programming languages like BASIC
8. Data Celler 12 2 Val V g mm Celler 12 3 Val Kraft N Kolumner B Kolumnbredd 8 43 End With With Diagramlista Diagram Data Diagramrubrik Titel V g Kraft Axlar xlKategori Axelrubrik Titel Axlar xlV rde Axelrubrik Titel End With With Diagramlista Diagram MedelVarde 53 Diagramrubrik Titel V g Kraft Diagram Axlar xlKategori Axelrubrik Titel Vag mm Axlar xlV rde Axelrubrik Titel Kraft N End With End With Case Vred With AktivArbetsbok With Kalkylbladlista Data Celler 10 2 Val Vinkel o Celler 10 3 Val Moment N cm Kolumner B Kolumnbredd 8 43 Kolumner C Kolumnbredd 14 End With With Kalkylbladlista Interpolerad Data Celler 10 2 Val Vinkel o Celler 10 3 Val Moment N cm Kolumner B Kolumnbredd 8 43 Kolumner C Kolumnbredd 14 End With With Kalkylbladlista Accumulerad Data Celler 12 2 Val Vinkel o Celler 12 3 Val Moment N cm Kolumner B Kolumnbredd 8 43 Kolumner C Kolumnbredd 14 End With With Kalkylbladlista MedelV rde Data Celler 12 2 Val Vinkel o Celler 12 3 Val Moment N cm Kolumner B Kolumnbredd 8 43 End With With Diagramlista Diagram Data Diagramrubrik Titel Vinkel Moment Diagram Axlar xlKategori Axelrubrik Titel Vinkel o Axlar xlV rde Axelrubrik Titel Moment N cm End With With Diagramlista Diagram MedelVarde
9. IETS f Ga Conversions The Force and Torque sensors are calibrated in Ounce Oz OzInch and radians rad In the robot code these units are converted to N Ncm and degrees as follows 1 N 23 596 Oz 1 Oz 28 35 g 0 2781 N 1 Nem 1 4156 OzInch 1 OzInch 0 7064 Ncm 1 rad 360 2z 57 296 degrees 65
10. In Section 2 1 the measurement situation will be described and in Section 2 2 some ideas of the desired measurement procedure are developed 2 1 Measurement Situation Volvo provided a set of different buttons switches and levers In order to narrow the scope of the work 4 push buttons and 3 rotary switches were selected for further tests A reasonable test cycle for the push buttons would be to push down on the button and at the same time measure the required force By recording the distance travelled and the equivalent required force at that particular distance the test result would consist of an array with distance force pairs The ideal velocity when pushing down on the button would probably be rather high to match real life operation but this will unfortunately demand a very fast sampling rate If the required force is a function of the velocity the result will of course be influenced Nevertheless since the purpose of the test is to establish some kind of measure of feeling that is comparable with other tests it was assumed that as long as all the tests were made using the same motion velocity the requirement for high velocity could be disregarded A conceivable test cycle for the rotary switches would in this case be to turn the rotary switch and at the same time measure the angle and the equivalent torque Here the test result would be an array of angle torque pairs The same reasoning concerning the motion velocity that was co
11. Sweden has a research department entirely designated for testing the electrical systems and user interfaces incorporated in the company s car fleet The user interface comprises all of the controls available to the driver and the passengers to operate the car Some examples are the steering wheel the pedals the gear lever and the abundance of electrical switches adjustment levers and knobs In an ordinary car today there are about 50 to 80 of these different controls Also included in the term user interface is the presentation system informing the driver about the status of the car and the environment This research department closely cooperates with the departments responsible for design and development of the equipment mentioned above In the process of developing new controls there has to be an interaction between the two departments to ensure that the products meet the required quality standards For instance that they work properly in various kinds of climates and environments during their entire life span and have a life expectancy that is adequate but also that the controls have the desired feeling The former properties are currently tested in the laboratories by intricate but not very flexible mechanical devices and in climate chambers and this is fairly simple and straightforward The testing of the latter property the feeling presents a far more difficult problem As of today the only way of testing the feeling of a control is to
12. connection The most basic mode of communication is the use of a terminal emulator utility that allows for direct low level communication with the control system The software also has the ability to manage the storage retrieval and downloading of programs locations and variables Furthermore it can control the execution of programs in the controller and it also logs events and errors and performs diagnostic tests of the user memory A text editor is used to write and edit the programs which later are downloaded to the controller There is a text editor included in RCWIN but any other text editor is sufficient As mentioned before the programs are written in RAPL 2 and will initially have the filename extension txt The programs are then tokenized i e the text code is converted to RAPL 2 mnemonic codes After it has been tokenized the filename will have the extension rob The program is thereafter downloaded individually or together with other programs variable files var or locations files loc to the controller Once the program is downloaded in the controller it can be started from either the teach pendant the front panel from RCWIN or from another program on the host computer via the DDE 19 interface DDE stands for Dynamic Data Exchange and is an interface enabling communication and data exchange between RCWIN and hence the 18 controller and other programs like for example process controlling
13. it was at all possible to isolate a few parameters that strongly influence the feeling of a control and if it furthermore existed a way of measuring these parameters the matter of ordering products from sub contractors and ensuring a uniform feeling would be much easier In the same way as Volvo today can specify the design function and quality of a product that is ordered from a sub contractor Volvo would also be able to specify the desired feeling of the product Once a product is ordered it would be the responsibility of the sub contractors that the product fulfills the specification of those parameters Furthermore it would be desirable if the test method were adaptable to different types of controls The mechanical devices that are used today for expected lifetime and function testing are constructed for a single purpose only and for a specific control When a new control type is to be subjected to testing a completely new mechanical test device has to be constructed This is very expensive and time consuming 1 2 Problem Definition and Objectives In the previous Section it was concluded that it would be desirable if some important parameters could be extracted from the term feeling and if there was an adaptable method to measure these parameters In the case of a push button switch for example two obvious parameters are the force and the depression distance required for operating it and in the case of a rotary switch two other natu
14. let a test panel consisting of a couple of persons operate it and report the feeling This of course will result in a very subjective statement that will highly depend on the test person Even the term feeling is very hard to pinpoint It could for example be the force required to activate a switch how the surface texture on the switch is experienced or if it conveys a sense of good quality to the user A rough definition of the term feeling could be the summarized impression the consumer experience when using the product Traditionally when a new product is developed a couple of prototypes are manufactured by a sub contractor and then tested manually by Volvo to determine if it has the desired feeling If not it will be sent back to the manufacturer and readjusted This process can sometimes be repeated several times and costs a lot of time money and efforts When the product eventually is accepted it will be mounted in a car together with several other controls and then yet another problem occurs Since each of the controls has been tested individually with subjective methods the different controls in a car might not have the same feeling This imposes a great problem because a car manufacturer wants the cars to constitute a specific concept Everything in the car should follow this concept If a car is intended to give the driver the impression of high quality then all the controls should also have a sensation of high quality to them If
15. reached the switches The obvious advantage of using a servo gripper however would be avoiding the necessity to change tools and also be able to test round buttons Figure 6 c The robot arm mounted on a table together with the switches 25 Figure 6 f The servo gripper 26 6 3 Software The software was created in VBA as described in Chapter 5 2 It was developed in an object oriented fashion to be easier to adapt to different measurement situations and future upgrades The graphical user interface is presented in Figures 6 g to 6 k The Swedish language was used in the user interface for labels and button names Nytt TestObjekt Figure 6 g The layout of the Main Menu with the button names in Swedish The user can 66 66 select Information Settings Measurement Show Data Show Diagram New Test Object Open Save Print or Quit Figure 6 h When selecting M tning Measurement in the main menu the following menu is presented Here the user can chose the number of measurements and enter the name of the test person 27 Figure 6 1 This menu is shown when selecting TestObjekt New Test Object in the main menu The desired robot test program and switch type can be chosen as well as the length of the interpolation Figure 6 j The menu for showing the result diagrams The user can choose between interpolated data or mean value data 28
16. should not be confused with the degrees of freedom The degrees of freedom also depend on the types of joints and how they are connected A robot arm may even have more than six joints but still have a maximum of six degrees of freedom or less The robot arm described in this text may for example be fitted on a track It would then still have six degrees of freedom but a larger working area Adding an extra joint also enables the arm to reach certain positions in 14 several different ways At the free end of the last link of the arm there is a tool flange where an end effector 2 can be attached This end effector can for example be a servo or a pneumatic gripper and in combination with these a force and torque transducer can be installed which will be treated later on in Section 4 1 Figure 3 b The locations of the six rotating joints in the robot arm EE BN Radius including 6 CM rA tool extension Fa 3 00 in A TS T6mm E P d 1 2 ER pe i Fe M 330 mm 13 0070 N P A is E I n p x 1194 mm 47 00 in R365 mnt 14 36 in LI R866 mm 34 00 in N d 180 ES PN moe R864 mm 94 00 ind jd fe We e Ge H 210 ws P NEN eel C XN X1 HEI Je o ger N wen ed N b p ps NE AN j Y DESCH Y uod NM T N Maximum Reach of ELEVATION Centre
17. u Step Kalkylbladlista Interpolerad Data Celler 10 6 Val 51 StepInv 1 u Step AntalUtfMat Kalkylbladlista Accumulerad Data Celler 10 2 Val 1 Kalkylbladlista Accumulerad Data Celler 10 2 Val AntalUtfMat Kalkylbladlista MedelVarde Data Celler 10 2 Val AntalUtfMat Kalkylbladlista HuvudMeny Celler 1 1 Val AntalUtfM t For i 1 To N 1 x1 Kalkylbladlista Data Celler i 10 2 Val x2 Kalkylbladlista Data Celler i 11 2 Val yl Kalkylbladlista Data Celler i 10 3 Val y2 Kalkylbladlista Data Celler i 11 3 Val If x1 lt gt x2 Then k y2 yl x2 1 Do While xx lt x2 xx x1 y1 Kalkylbladlista Interpolerad Data Celler xx StepInv 11 2 Val xx Kalkylbladlista Interpolerad Data Celler xx StepInv 11 3 Val yy AntalMat Kalkylbladlista Accumulerad Data Celler xx StepInv 13 4 Val 1 Kalkylbladlista Accumulerad Data Celler xx StepInv 13 4 Val AntalMat gyy Kalkylbladlista Accumulerad Data Celler xx StepInv 13 3 Val Kalkylbladlista Accumulerad Data Celler xx StepInv 13 2 Val xx Kalkylbladlista Accumulerad Data Celler xx StepInv 13 3 Val yy gyy Kalkylbladlista MedelVarde Data Celler xx StepInv 14 2 Val xx Kalkylbladlista MedelVarde Data Celler xx StepInv 14 3 Val yy gyy AntalMat If AntalMat 1 Then Ka
18. 12 and Pascal 13 It allows among other things for robot control computations storage of variables and locations logical operations supervision of inputs and outputs and handling of subroutines There are two methods to implement a program in the controller By using the teach pendant one programming line at the time can be entered or edited When implementing larger and more complex programs however it is better to edit the program in Windows on a host computer and then download the complete program to the controller The procedures for writing and downloading programs are described in the next Section Another method for designing programs are the use of Process Control Programming PCP 4 14 which utilizes high level programming e g C C 15 and allows for a more efficient application of the controller system 17 3 4 Host Computer Supervisory amp Host computer 5j RCWIN Robcomm Figure 3 f The host computer consists of an ordinary IBM compatible PC in this case with a Pentium processor and the operative system Windows 95 The host computer Figure 3 f consists of an ordinary IBM compatible PC The computer used in these tests has a Pentium 16 processor and is using the operative system Windows 95 17 Installed on the computer is the robot communication software Robcomm for Windows RCWIN 18 This software enables the communication between the PC and the C500 robot controller using a 38400 bps serial
19. 5 measurements 7 2 Analysis Push Buttons The figures above show that each push button has a rather distinct profile In some cases the required force rises to a near constant value until the button is fully pressed down Figure 7 c For other buttons a higher initial force is required Figure 7 a Analysis of data from one measurement Figure 7 c to another one 2 days later Figure 7 e using the Pearson Product Moment Correlation 24 in Excel Appendix C gives a correlation coefficient of 0 98 Same analysis but using mean values Figure 7 d and Figure 7 f gives a correlation coefficient of 0 99 This suggests that the tests are repeatable over time Another interesting property is the energy content of the button How much energy is used to press down the button Since the travel mm and force N are measured the area under the graph converted to standard scientific units should represent that energy Using the trapezoidal numerical method 25 Appendix C the energy content for pressing the buttons were calculated and presented in the figures above A comparison of the energy levels between several tests on the same button also shows that the tests are repeatable and that this property can be used to define the button profile The maximum and minimum values presented in the figures above indicate that there is not much variation in the measured results from consecutive tests 36 7 3 Test Results Rotary Swit
20. BCOMM Topic Tex RUN Item Tex Main On Error GoTo 150 DDEKanal DDEInitiera Application Topic DDEUtf r DDEKanal Item DDEAvsluta DDEKanal Exit Sub 150 MsgBox Ett fel har uppstatt Prova igen VisaHuvudMeny End Sub 50 Sub DDES ttVariabel Application VarName VarValue Subrutinen ppnar en DDE f rbindelse med programmet Application s tter en variabel VarName till v rdet VarValue och st nger d refter DDE f rbindelsen Application Tex ROBCOMM VarName Ett variabelnamn VarValue V rdet p variabeln sTmpString For nCounter 1 To 8 Len sVarName sTmpString sTmpString amp Chr 32 Next nCounter Item CStr VarName amp TmpString amp amp VarValue On Error GoTo 150 DDEKanal DDEInitiera Application VARIABLE DDEPoke DDEKanal Item DDEAvsluta DDEKanal Exit Sub 150 MsgBox Ett fel har uppstatt Prova igen VisaHuvudMeny End Sub Sub Inst llningar InstallRuta Indataruta Ange l senord Inst llningar Om Inst llRuta crs S Init Slut Sub Sub F rdr jning i VantaTid sekunder KlarTid Timer VantaTid If KlarTid gt 86400 Then KlarTid KlarTid 86400 Do Until Timer gt KlarTid Till t systemet att utf ra andra uppgifter DoEvents Loop End Sub Sub Sampla N Interpolerar data Antal datapunkter N f r kontakt nr TestObjekt Step ar den nskade stegl ngden O
21. CODEN LUTEDX TEIE 5353 1 65 2015 Flexible Product Testing Using an Articulated Robot System Equipped with a Real Time Force and Torque Sensor C H JUS jy S dum 2 lt v e d d EN be U ja U D EN Ulf Tronde Division of Industrial Electrical Engineering and Automation Faculty of Engineering Lund University Industrial Flexible product testing using an articulated robot system eguipped with a real time force and torque sensor Master Thesis by UIf Tronde Division of Industrial Electrical Engineering and Automation IEA Supervisor IEA Gunnar Lindstedt Supervisor CRS Robotics G ran Lasson Examinator IEA UIf Jeppsson Faculty of Engineering Lund University May 2015 Abstract Car manufacturers have traditionally tested new products using intricate but not very flexible mechanical devices The test methods have been designed to check that e g the user interfaces and controls work properly in various kinds of climates and environments and have a life expectancy that is adeguate It has however been difficult to test that the controls have the desired feeling As of today the only way of testing the feeling of a control is to let a test panel consisting of a couple of persons operate it and report the feeling This will result in a very subjective statement that will highly depend on the test persons The term feeling is difficult to define b
22. Data Celler 4 2 Val Kalkylbladlista Data Celler 4 2 Val Kalkylbladlista MedelVarde Data Celler 5 2 Val Kalkylbladlista Data Celler 5 2 Val Kalkylbladlista MedelVarde Data Celler 6 2 Val Kalkylbladlista Data Celler 6 2 Val Kalkylbladlista MedelVarde Data Celler 10 2 Val 0 End Sub Sub NyttTestObjektAvbryt Knapp Dialogbladlista Dialog3 D61j AktivArbetsbok Bladlista HuvudMeny Aktivera End Sub Sub ppna Knapp filAttOppna Program HamtaOppnaFilnamn Excel Ck xls xls If filAttOppna lt gt False Then Arbetsb cker ppna filnamn filAtt ppna End If End Sub Sub Spara Knapp GammalS kv g C crs userprog prog_1 excel GammaltFilNamn Kalkylbladlista Data Celler 3 2 Val 56 AktivArbetsbok SparaKopiaSom GammalS kv g GammaltFilNamn XLS End Sub Sub SkrivUt Knapp Ej aktiverad Printfunktion finns inkorporerad i Excel End Sub Sub Avsluta Knapp End End Sub Sub M tning Knapp Dialogbladlista Dialog4 Visa End Sub Sub M tningOK Knapp Kalkylbladlista Data Celler 7 2 Val Dialogbladlista Dialog4 Redigeringsrutor 8 u Text Kalkylbladlista Interpolerad Data Celler 7 2 Val Kalkylbladlista Data Celler 7 2 Val Dialogbladlista Dialog4 D61j Matning CInt Dialogbladlista Dialog4 Redigeringsrutor 10 u Text End Sub Sub M tningAvbryt Knapp Dialogbladlista Dialog4 D lj End Sub Sub VisaData Kn
23. FORCE STARTF gt MAXCONOZ THEN 100 C UNTIL C gt 100 FINISH TREO AR AACR oce IRC KK COUNT 1 PASSWORD 255 CLINSPD MAXLNSPD SENSOR FORCE CALIBRATE X STARTF FORCE 9 SPEED SPEED1 Z DEPTH1 S 10 ACTUAL LOCAT COUNT KRAFT COUNT FORCE 0 STARTF IF ABS FORCE STARTF gt MAXFOZ THEN 20 COUNT GOTO 10 20 HALT KRAFT COUNT 1 9999 BUSY 0 CLINSPD 40 0 PASSWORD 0 SPEED 30 MOVE PBEnd Object FINISH SPEED 50 MOVE PBStart Object FINISH GOTO 200 100 HALT SPEED 50 MOVE PBStart Object FINISH 200 BUSY 0 IF COUNT MAXNUMB THEN 999 STOP 999 MAXNUMB COUNT STOP 62 RotarySwitch TXT Program Name RotarySwitch This program does the actual measurement of a rotary switch designated Object The operator have to define start positions Rwstarti1 0bject RWStart2 Object and an end position RSEnd Object for each test object BUSY 1 POSE FUN GOSUB FON SENSOR FORCE FRAME RAW MAXFOZ 3 596 MAXFN3 Converts from N to Oz I MAXCONOZ 3 596 MAXCONN Converts from N to Oz VINKE12R VINKE12 57 29578 Conv degrees to rad SPEED 50 MOVE RWStart1 Object Move to start position MOVE RWStart2 Object FINISH SENSOR FORCE CALIBRATE X STARTF FORCE 0 SPEED 10 MOVE RSEnd Object Start turning the switch IR RR ERE ERE EE Ke SAFETY Interrupts if anything in the path if forc
24. HamtaDDETal ROBCOMM LOCARRAY LOCAT KRAFT HamtaDDETal ROBCOMM VARARRAY KRAFT VinkelStart 57 29578 x KonvDDELocArrayTillDbl LOCAT 1 1 6 KRAFTSTART KonvDDEVarArrayTillDbl KRAFT 1 1 Kalkylbladlista Data Omrade Area 11 2 500 3 RaderaInneh ll N 1 Do While KonvDDEVarArrayTillDbl KRAFT N 1 lt 9990 Vinkel 57 29578 x KonvDDELocArrayTillDbl LOCAT N 1 6 If Vinkel VinkelStart 0 2 Then Vinkel Vinkel 360 ElseIf Vinkel VinkelStart gt 360 Then Vinkel Vinkel 359 9 End If Kalkylbladlista Data Celler N 10 2 Val Vinkel VinkelStart 359 9 48 Kalkylbladlista Data Celler N 10 3 Val 0 706408 x KonvDDEVarArrayTillDbl KRAFT N 1 KRAFTSTART N N 1 Loop Sampla N 1 Exit Sub 110 MsgBox Ett fel har uppst tt Prova igen 110 VisaHuvudMeny End Sub Function KonvDDEVarArrayTillDbl DDETal Konverterar variabel fr n str ng till dubbelvariable On Error GoTo 120 A InStr 1 DDETal 0 u_b Len DDETal DDETalxx Right DDETal u_b A C InStr 1 DDETalxx 9 Mid DDETalxx 1 KonvDDEVarArrayTillDbl CDbl DDETalxx Exit Function 120 MsgBox Ett fel har uppst tt Prova igen 120 VisaHuvudMeny End Function Function KonvDDELocArrayTillDbl DDETal AntalKomma Konverterar variabel fr n str ng till dubbelvariable On Error GoTo 121 aa 1 For i 1 To AntalKomma A InStr aa DDETal 0 aa 1 Nex
25. and database software In these tests the DDE interface is used to connect RCWIN to Microsoft Excel and Visual Basic for the purpose of launching the test programs and processing and storing the measured data provided by the controller and the force and torgue sensing eguipment see Chapter 5 for a more detailed description of this issue 19 4 The Real Time Force and Torgue Sensor System The force and torgue sensor system 20 21 is manufactured by ATI Industrial Automation North Carolina USA The system operates in two different modes force sampling mode and force control mode In force sampling mode the force sensor provides information to the robot controller but plays no part in controlling the action of the robot This is the operating mode that was used in this thesis In force control mode the sensor provides information to the control loop regulating the robot arm This allows for adaptive control of the forces the arm is exerting on surrounding objects The system consists of the transducer see Section 4 1 which is mounted on the robot arm and does the actual force and torgue sampling and the F T controller see Section 4 2 that process the sampled signals and communicates with the robot controller 4 1 The Transducer Force Tordue Transducer Figure d a The Force Torgues Transducer at the end of the robot arm Figure 4 b A closer look at the Gamma 15 50 transducer 20 The sensor part of the Gamma
26. app Dialogbladlista Dialog1 Visa End Sub Sub VisaDiagram Knapp Dialogbladlista Dialog2 Visa End Sub 57 Sub OriginalData Knapp Dialogbladlista Dialog1 D lj AktivArbetsbok Kalkylbladlista Data Aktivera End Sub Sub InterpoleradData Knapp Dialogbladlista Dialog1 D61j AktivArbetsbok Kalkylbladlista Interpolerad Data Aktivera End Sub Sub Medelv rdesbehandladData Knapp Dialogbladlista Dialog1 D lj AktivArbetsbok Kalkylbladlista MedelV rde Data Aktivera End Sub Sub VisaDataAvbryt Knapp Dialogbladlista Dialog1 D lj AktivArbetsbok Bladlista HuvudMeny Aktivera End Sub Sub Information Knapp Ej aktiverad End Sub Sub Inst llningar Knapp Ej aktiverad End Sub Sub VisaDiagramData Dialogbladlista Dialog2 D61j AktivArbetsbok Bladlista Diagram Data Aktivera End Sub Sub VisaDiagramMVData Dialogbladlista Dialog2 D61j 58 AktivArbetsbok Bladlista Diagram MedelVarde Data Aktivera End Sub Sub VisaDiagramAvbryt Knapp Dialogbladlista Dialog2 D61j AktivArbetsbok Bladlista HuvudMeny Aktivera End Sub Sub VisaHuvudMeny AktivArbetsbok Bladlista HuvudMeny Aktivera End Sub 59 Appendix B Robot Files Robot Code RAPL 2 Code executed in Robcomm 4 30 INIT TXT Program Name Init Initializes the robot system and sets the speed BUSY 1 POSE FUN PASSWORD 255 CLINSPD 40 0 PASSWORD 0 SPEED 30 READY FINISH
27. ches In the diagrams below the angle degrees and the correlating measured torque Ncm are depicted on the x and the y axes The Swedish terms Vinkel Moment Diagram Vinkel and Moment can be translated to Angle Torque Diagram Angle and Torque Vinkel Moment Diagram Moment N cm Vinkel o Figure 7 k Rotary switch 1 I measurement First test Energy 51 2 mJ Vinkel Moment Diagram Min Max Moment N cm Vinkel Figure 7 1 Rotary switch 1 Max Mean Min values of 100 measurements First test Vinkel Moment Diagram Moment N cm Vinkel o Figure 7 m Rotary switch 1 1 measurement Second test 3 weeks after first test Energy 47 6 mJ Vinkel Moment Diagram Medel Min Moment N cm Vinkel o Figure 7 n Rotary switch 1 Max Mean Min values of 100 measurements Second test 3 weeks after first test Vinkel Moment Diagram 7 000 5 000 Vinkel o Figure 7 0 Rotary switch 2 I measurement Energy 37 8 mJ Vinkel Moment Diagram Moment N cm Vinkel o Figure 7 p Rotary switch 2 Max Mean Min values of 5 measurements Min Max Vinkel Moment Diagram 10 000 Vinkel o Figure 7 q Rotary switch 3 I measurement Energy 92 0 mJ 40 7 4 Analysis Rotary Switches Just as for the push buttons the fi
28. ction Analog to Digital conversion Means equipped with joints American Standard Code for Information Interchange The code is used in computer communication and represents text It encodes 128 specific characters Bits per second Central Processing Unit Dynamic Data Exchange Protocol used for communication and data sharing between applications in Windows Degrees of Freedom Non volatile memory Contains firmware and settings that remains also after power is removed Liquid Crystal Display A connector with a built in light emitter and receiver which makes it galvanically isolated Personal Computer Proportional Integral Derivative An algorithm that uses three terms consisting of the proportional the integral and the derivative values to control a process PLC Repeatability Pose repeatability RISC RS232 VB VBA Programmable Logic Controller A computer or application specifically designed for controlling automation Random Access Memory A volatile memory where data can be read and written with high speed The data is lost when power is removed Closeness of agreement between the positions and orientations of the attained poses after n repeated visits to the same commanded pose in the same direction Reduced Instruction Set Computing A microprocessor architecture where the number of instructions are reduced to improve processing speed A serial communication standard Vis
29. datering False DDEKommando ROBCOMM RUN RobotProgNamn Wait 30 Do While RobotBusy 1 Loop Select Case Typ Case TryckKnapp HamtaDataXKraft Case Vred HamtaDataFiMoment End Select Program Skarmuppdatering True Next Avsluta Sub 50 MedRuta Ett fel har uppst tt Prova igen VisaHuvudMeny End Sub Function RobotBusy On Error GoTo 80 RobotBusyDDE HamtaDDETal ROBCOMM VARIABLE BUSY RobotBusy KonvDDEVariableTillHeltal RobotBusyDDE 1 1 Exit Function 80 MsgBox Ett fel har uppstatt Prova igen VisaHuvudMeny End Function 47 Sub H mtaDataXKraft H mtar variabel arrayerna f r kraft och position Vid Fel G Till 90 LOCAT HamtaDDETal ROBCOMM LOCARRAY LOCAT KRAFT H mtaDDETal ROBCOMM VARARRAY KRAFT LOCATSTART KonvDDELocArrayTillDbl LOCAT 1 1 2 KRAFTSTART KonvDDEVarArrayTillDbl KRAFT 1 1 Kalkylbladlista Data Omr de Area 11 2 500 3 RaderaInnehall N 1 Do While KonvDDEVarArrayTillDbl KRAFT N 1 lt 9990 Kalkylbladlista Data Celler N 10 2 Val 1 KonvDDELocArrayTillDbl LOCAT N 1 2 LOCATSTART Kalkylbladlista Data Celler N 10 3 Val 0 27811 KonvDDEVarArrayTillDbl KRAFT N 1 KRAFTSTART N N 4 1 Loop Sampla N 1 Avsluta Sub 90 MedRuta Ett fel har uppstatt Prova igen VisaHuvudMeny End Sub Sub HamtaDataFiMoment H mtar variabel arrayerna for vridmoment och position On Error GoTo 110 LOCAT
30. e more than max allowed force 1 C 0 DO IF ABS FORCE 0 STARTF gt MAXCONOZ THEN 100 C UNTIL C gt 50 FINISH TR RR IK KIKKER Ke COUNT 1 SPEED 50 ROLL VINKE11 FINISH STARTF FORCE 3 ACTUAL STARTL A TCOMP STARTL VINKSLUT TCOMP STARTL VINKE12R IF VINKSLUT lt 3 141592 THEN 300 IF VINKSLUT 3 141592 THEN 350 5 SPEED SPEED3 ROLL VINKE12 DO ACTUAL LOCAT COUNT HAR TCOMP LOCAT COUNT KRAFT COUNT FORCE 3 STARTF IF ABS FORCE 3 STARTF MAXFOZ THEN 20 COUNT 63 UNTIL ABS HAR VINKSLUT lt 0 02 KRAFT COUNT 9999 BUSY 0 SPEED 50 ROLL VINKE11 VINKE12 VINKE13 ROLL VINKE13 MOVE RWStart1 0bject FINISH IF COUNT MAXNUMB THEN 999 STOP 20 HALT KRAFT COUNT 1 9999 BUSY 0 SPEED 50 MOVE RSEnd 0bject FINISH ROLL VINKE13 ROLL VINKE13 MOVE RWStart1 0bject FINISH IF COUNT MAXNUMB THEN 999 STOP 100 HALT SPEED 50 MOVE RWStart1 0bject FINISH STOP 300 VINKSLUT VINKSLUT 6 28318 GOTO 5 350 VINKSLUT VINKSLUT 6 28318 GOTO 5 999 MAXNUMB COUNT STOP 64 Appendix C Formulas and Conversions Formulas The Pearson Product Moment Correlation Coefficient 24 for two sets of values calculated using the formula Correlation in Excel _ _ X 36 39 Xty y The Trapezoidal numerical method 25 for calculating the area under the graph oa 15 i
31. ectives over Time The Master Thesis work was performed late 1996 to early 1997 which was 18 years ago At that time the IT sector had just started for real and since that time there has been a tremendous development and also the introduction of completely new platforms The car industry has of course also had a large development but compared to the IT sector it has been modest In the car industry there is still a demand for user tests like the one developed in this work The interior of the car contains a majority of the user interfaces that existed 18 years ago The introduction of touch screens is the major exception and will probably revolutionize the industry in the next years to come Nonetheless the problem definition and approach in this thesis is still very much valid The computer platforms and software used in this work have certainly evolved but all of them are still available in new versions and are widely used The same is true for the robot system Below is a list of the different components and the status of today The small laboratory and industrial robots from CRS Robotics are available in newer versions from Thermo Fisher Scientific The specifications for their robot F5 are very similar to that of the A465 with improved repeatability and strength The software controlling the robot is more user friendly e The force and torque sensors are still available from the same manufacturer e Personal computers PC using Mic
32. g but also to measure force and torgue in real time during these tasks 1 4 Overview of the Thesis Chapter 1 provides the background for the study and in Chapter 2 the problem of finding reasonable measurement methods for different types of controls is discussed together with a description of the desired measurement procedure An outline of the robot system both hardware and software is found in Chapter 3 In Chapter 4 the force and torque sensor is described and Chapter 5 treats the supervisory computer system The entire integration of the equipment and the test procedure is covered in Chapter 6 and in Chapter 7 the results are presented and analyzed and the chapter also contains the conclusions In Chapter 8 the study is discussed providing topics for future works and specific challenges encountered The programming code is attached in Appendices A and B and formulas and conversions are found in Appendix C 11 2 Problem Approach In Section 1 2 it was pointed out that the two basic operations when interacting with the car controls are pushing or pulling and rotation Examples of the former are activation of an electrical pushbutton changing the gear and maneuvering the windshield wiper stick Examples of the latter are turning of knobs and turning the steering wheel In order to limit the extent of the thesis a couple of ordinary electrical push buttons and rotary switches were selected to represent the various types of controls
33. gures above show that the rotary switches also have rather distinct profiles It can be clearly seen that the graphs correspond to the notches that can be felt when turning the rotary switches Analysis of data from one measurement Figure 7 k to another one 3 weeks later Figure 7 m using the Pearson Product Moment Correlation 24 in Excel Appendix C gives a correlation coefficient of 0 94 Same analysis but using mean values Figure 7 1 and Figure 7 n gives a correlation coefficient of 0 98 These tests also seem to be repeatable over time The energy required to turn the rotary switch should also be possible to calculate With the angle in radians and torgue Ncm measured the area under the graph converted to standard scientific units should represent the energy Using the trapezoidal numerical method 23 Appendix C the energy content for the rotary switches were calculated and presented in the figures above Also here a comparison of the energy levels between several tests on the same switch shows that the tests are repeatable and that this property can be used to define the switch profile The maximum and minimum values presented in the figures above indicate that there is not much variation in the measured results from consecutive tests 7 5 Conclusions From the performed tests and result analysis it can be concluded that e 115 possible to measure the test objects as desired with a standardized method using the ar
34. he buttons typically had a travel distance of 5 7 mm Figure 6 a The push buttons used for the tests Labelled 1 to 5 from left to right The rotary switches used in the test Figure 6 b were integrated on a climate control panel that was fixed on top of a table The panel was mounted with an inclination of about 30 relative to the table The rotary switches were labelled 1 to 3 starting with the one to the left Switches 1 and 3 had a full travel of 180 from position 9 o clock to 3 o clock and o clock to 12 o clock The switch in the middle number 2 had no travel limitations and could thus be rotated continuously to the left or right All three switches had notches that could be felt when turning them 24 Figure 6 b The rotary switches used for the tests Labelled 1 to 3 from left to right 6 2 Hardware The robot arm was mounted on a table together with the switches Figure 6 c At the free end of the last link of the arm a force and torque transducer was fitted to the tool flange Then in turn either a pin tool Figure 6 d or a fork shaped tool Figure 6 e was attached to the transducer The pin tool was used to press the push button switches and the fork shaped tool was used to turn the rotary switches Some trials were initially made using a servo gripper Figure 6 f but it appeared to be somewhat heavy and cumbersome The pin tool and the fork shaped tool however were light and easily
35. it Initierar robotsystemet Vid Fel GaTill 10 DDEKommando ROBCOMM RUN Init Initierad Sant Avsluta Sub 10 MedRuta Ett fel har uppstatt Prova igen VisaHuvudMeny Slut Sub Sub Matning AntalMatningar Kors nar knappen Matning aktiveras Forbereder kalkylblad och k r sedan robotens testsekvens Namn RobotProgNamn pa testobjekt nr Objekt Detta utf rs AntalMatningar ggr H mtar sedan Kraft eller Vridmoment arrayvariabel Vid Fel G Till 50 RobotProgNamn Kalkylbladlista Interpolerad Data Celler 11 6 Val Object Kalkylbladlista Interpolerad Data Celler 12 6 Val Typ Kalkylbladlista Data Celler 4 2 Val Kalkylbladlista Data Celler 8 2 Val Now 46 Kalkylbladlista Interpolerad Data Celler 8 2 Val Kalkylbladlista Data Celler 8 2 Val Kalkylbladlista Accumulerad Data Celler 8 2 Val Kalkylbladlista Data Celler 8 2 Val Kalkylbladlista MedelVarde Data Celler 8 2 Val Kalkylbladlista Data Celler 8 2 Val If Kalkylbladlista Accumulerad Data Celler 10 2 Val 0 Then Kalkylbladlista Accumulerad Data Celler 7 2 Val Kalkylbladlista Data Celler 8 2 Val Kalkylbladlista MedelVarde Data Celler 7 2 Val Kalkylbladlista Data Celler 8 2 Val End If Om Initierad Falskt S Init V nta 5 Utf r Medan RobotBusy 1 Loop Slut Om DDES ttVariabel ROBCOMM Object Object For i 1 To AntalM tningar Program Sk rmupp
36. ladlista Accumulerad Data Omr de Area 7 2 8 2 RaderaInnehall Kalkylbladlista MedelVarde Data Omr de Area 14 2 500 5 RaderaInnehall Kalkylbladlista MedelVarde Data Omr de Area 7 2 8 2 RaderaInnehall Kalkylbladlista Interpolerad Data Celler 3 2 Val Kalkylbladlista Data Celler 3 2 Val Kalkylbladlista Interpolerad Data Celler 4 2 Val Kalkylbladlista Data Celler 4 2 Val Kalkylbladlista Interpolerad Data Celler 5 2 Val Kalkylbladlista Data Celler 5 2 Val Kalkylbladlista Interpolerad Data Celler 6 2 Val Kalkylbladlista Data Celler 6 2 Val 55 Kalkylbladlista Interpolerad Data Celler 19 6 Val Kalkylbladlista Info Celler Dialogbladlista Dialog3 NedrullbaraListrutor 32 Val 5 5 Val Kalkylbladlista Interpolerad Data Celler 11 6 Val Dialogbladlista Dialog3 Redigeringsrutor 37 u Tex t Kalkylbladlista Accumulerad Data Celler 3 2 Val Kalkylbladlista Data Celler 3 2 Val Kalkylbladlista Accumulerad Data Celler 4 2 Val Kalkylbladlista Data Celler 4 2 Val Kalkylbladlista Accumulerad Data Celler 5 2 Val Kalkylbladlista Data Celler 5 2 Val Kalkylbladlista Accumulerad Data Celler 6 2 Val Kalkylbladlista Data Celler 6 2 Val Kalkylbladlista Accumulerad Data Celler 10 2 Val 0 Kalkylbladlista MedelVarde Data Celler 3 2 Val Kalkylbladlista Data Celler 3 2 Val Kalkylbladlista MedelVarde
37. lkylbladlista MedelVarde Data Celler xx StepInv 14 Kol 4 Val yy Kalky lbladlista MedelVarde Data Celler xx StepInv 14 Kol 5 Val yy End If If yy lt Kalkylbladlista MedelVarde Data Celler xx StepInv 14 4 Val Then Kalkylbladlista MedelVarde Data Celler xx x StepInv 14 Kol 4 Val yy If yy gt Kalkylbladlista MedelVarde Data Celler xx StepInv 14 5 Val Then Kalkylbladlista MedelVarde Data Celler xx x StepInv 14 Kol 5 Val yy 52 End Next End Sub XX XX u Step Loop If Function Area Rad V Kol V RadUH KolUH Area Chr Kol V 64 CStr Rad V i Chr KolUH 64 CStr RadUH End Function Sub InitTyp Typ F rbereder kalkylbladslistor och diagram Select Case Typ Case TryckKnapp Step With AktivArbetsbok Diagram V g mm Kraft N Data With Kalkylbladlista Data Celler 10 2 Val Vag mm Celler 10 3 Val Kraft N Kolumner B Kolumnbredd 8 43 Kolumner C Kolumnbredd 8 43 End With With Kalkylbladlista Interpolerad Data Celler 10 2 Val V g mm Celler 10 3 Val Kraft N Kolumner B Kolumnbredd 8 43 Kolumner C Kolumnbredd 8 43 End With With Kalkylbladlista Accumulerad Data Celler 12 2 Val V g mm Celler 12 3 Val Kraft N Kolumner B Kolumnbredd 8 43 Kolumner C Kolumnbredd 8 43 End With With Kalkylbladlista MedelVarde
38. n the stand alone software development tool Visual Basic VB The difference is that programs developed in VB are detached and can be used without any other programs while programs developed in VBA included in Excel is closely connected to this program and has to be run together with it which is the case in this work Excel and VBA was chosen because it is reliable well known and easy to maintain and migrate It is possible for anyone with some programming knowledge to edit and make changes in the program to adapt it to a specific measurement situation Another feature of Excel and VB VBA is the use of DDE for communication with other applications as mentioned in Section 3 4 23 6 Test Implementation This chapter provides an overview of the complete testing system 6 1 Test Objects The selected push buttons Figure 6 a were mounted facing upwards on an aluminum panel that was fixed on top of a table They are referred to as button numbers 1 to 5 starting with the one to the left Button 1 had an integrated rheostat wheel Trying to measure this turning movement seemed a difficult task for this initial system so button 1 was therefore excluded from the tests Button 2 was of the seesaw type and was turned on by pushing down on one end and turned off by pushing down on the other end Buttons 3 to 5 were activated by pushing down once and deactivated by pushing down a second time Only the activation phase was tested T
39. nducted for the push buttons is assumed to be valid also for rotary switches 12 2 2 Desired Measurement Procedure The desired system should be able to perform safe and reproducible tests with minimal interference from the operator By selecting a specific type of test and placing the test object in a predetermined position and then activating the system it should by itself perform one or more test cycles and then process the result and finally present it to the operator A brief summary of the desired elements included in the measurement procedure is as follows performed by the operator performed by the system Place test object in a predetermined position Select type of measurement push button or rotary switch Option Select number of cycles motion velocity etc Start the test Execute the test cycle s and sample distance force or angle torque Perform necessary signal processing and conversions Save the result in a computer file Process the data compute averages and statistics etc Present the result Select desired presentation form data diagram etc Save print or compare test objects e Chose to continue testing or to shut down 13 3 The Robot System The robot system was provided by CRS Robotics Nordic AB Lund Sweden Head office CRS Robotics Corporation Burlington Ontario Canada Current name Thermo Fisher Scientific Waltham Massachusetts USA The system Figure
40. of Wrist Pitch Axis J5 ee Figure 3 c The working area is the space that is within reach for the tool tip This particular robot arm is capable of manipulating a nominal payload of 2 kg and a maximum payload of 3 kg It has a repeatability of 0 05 mm This means that it is able to return to a specific point with an absolute error less than 0 05 mm even after several cycles Repeatability should not be mixed up with accuracy 3 Accuracy is the deviation between the desired computed position and the actual attained position There is no accuracy stated for this robot but generally it can be assumed that the repeatability is better than the accuracy Furthermore the arm has the ability to move at a maximum velocity of 4 5 m s 1 When performing linear or path motions the maximum velocity is reduced to 1 m s The arm joints are driven by DC servo motors via harmonic drives which provides gear reduction for higher torque The position of each joint is determined with optical encoders 15 3 2 The Robot Controller Figure 3 d The C500 robot controller from CRS Robotics Corporation The C500 robot controller unit Figure 3 d is the electronic heart of the robot system 4 Based on a 32 bit RISC 5 processor it performs the necessary calculations to control the motions of the robot arm It is able to simultaneously support up to eight robot axis and a gripper with an update rate of 14 kHz corresponding to an update time of 70 u
41. raft N V g mm Figure 7 b Push Button 2 Max Mean Min values of 5 measurements Vag Kraft Diagram 14 000 12 000 10 000 2 000 2 000 V g mm Figure 7 c Push Button 3 I measurement first test Energy 15 5 mJ Vag Kraft Diagram 16 000 14 000 12 000 10 000 500 Medel Min Max Kraft N 4 000 2 000 0 000 2 000 V g mm Figure 7 d Push Button 3 Max Mean Min values of 200 measurements first test Vag Kraft Diagram 14 000 12 000 10 000 8 000 6 000 Kraft N 4 000 2 000 2 000 V g mm Figure 7 e Push Button 3 I measurement second test 2 days after first test Energy 15 9 mJ Vag Kraft Diagram 16 000 14 000 12 000 10 000 Medel Max Kraft N g 3 s 4 000 2 000 2 000 V g mm Figure 7 f Push Button 3 Max Mean Min values of 50 measurements second test 2 days after first test Vag Kraft Diagram Kraft N V g mm Figure 7 g Push Button 4 I measurement Energy 8 64 mJ Vag Kraft Diagram Medel Min Kraft N Vag mm Figure 7 h Push Button 4 Max Mean Min values of 50 measurements Vag Kraft Diagram V g mm Figure 7 i Push Button 5 I measurement Energy 7 10 mJ Vag Kraft Diagram Min Max V g mm Figure 7 j Push Button 5 Max Mean Min values of
42. ral parameters are the torque and the rotation angle When studying other types of controls it seems as if these particular parameter pairs form the basis In consideration of the desire to work out a flexible test method for measuring force torque and adherent positions and angles it appears to be a natural step to use a small industrial robot system that has the ability to do these types of measurements The main objective of this master thesis is to develop and evaluate a suitable and standardized method to measure force and torque during product testing using an articulated robot system equipped with a force and torque sensor 1 3 Flexible Systems for Product Testing The articulated robot system from CRS Robotics Corporation is designed for a diversified field of applications Since it is easily reprogrammed and reconfigured it provides a very flexible method to automate industrial tasks such as in this case product testing Currently the customary method to test products is to use equipment developed for one particular type of test and often these tests have to be performed manually By instead using a robot system a much more adaptable method can be derived and by eliminating the need for manual manipulation the tests should be more accurate and repetitive When the robot system is equipped with a force and torque sensor it has the ability to not only perform traditional product testing like function and life 10 time testin
43. ration Burlington Ontario Canada 1996 15 http www iso org iso home store catalogue_tc catalogue detail htm csnumber 64029 16 http www intel se content www se sv processors pentium pentium processor html 17 https support microsoft com en us kb 138349 18 Robcomm User s Guide UMI 23 504 CRS Robotics Corporation Burlington Ontario Canada 1996 44 19 20 21 22 23 24 25 https msdn microsoft com en us library windows desktop ms648774 v vs 85 aspx Installation and operations manual for F T PN 9610 05 1001 11 ATI Industrial Automation Apex North Carolina USA 1994 Force Sensor User s Guide UMI 23 510 CRS Robotics Corporation Burlington Ontario Canada 1996 http www ascii code com https msdn microsoft com en us library office ee814737 v office 14 aspx http www excelfunctions net Excel Correl Function html Torgil Ekman Numeriska Metoder p Dator och Dosa ISBN 9144262418 Studentlitteratur Lund Sweden 1987 All websites were reachable at 10 30 CET 2015 05 20 45 Appendix A Visual Basic Applications VBA Code VBA version 5 0 included in Microsoft Excel version 95 Swedish edition Allm n Initierad Som Bool Sub Auto ppna Definierar s kv gar RobcommS kv g C crs rcwin ChDir RobcommSokvag Shell RobcommSokvag robot exe 6 AppActivate Microsoft Excel End Sub Sub Auto_stang Initierad Falskt End Sub Sub In
44. rosoft Windows are still in use just much more powerful Excel and the included VBA macros are still widely spread and used and DDE is also still used within Microsoft Windows and Microsoft applications The VBA code would have to be adjusted for new syntax codes and communication protocols In summary it is my belief that this Master Thesis work still is of interest today with rather much the same problem definition and approach as well as result Another aspect of course is the reflection that it would just have been so much easier to finalize and present the report 18 years ago UIf Tronde 21 May 2015 Acknowledgments I would like to thank my supervisors at the Division of Industrial Electrical Engineering and Automation IEA Gunnar Lindstedt and Gustaf Olsson my supervisor at CRS Robotics G ran Lasson and my examinator at IEA UIf Jeppsson Also thanks to Ulf Edvardsson Volvo Product amp Process Engineering Division for supplying test specimens and helping to define the background of the work My greatest gratitude though goes to my wife Ann who has been supporting and encouraged me to finally complete this master thesis report Abbreviations and Terms Accuracy Pose Accuracy A D converter Articulated ASCII bps CPU DDE DOF FLASH EPROM LCD Opto isolated connector PC PID Deviation between the commanded pose and the attained pose when approaching the commanded pose from the same dire
45. s per axis It uses a PID 6 regulator control algorithm and has either a trapezoidal or parabolic velocity profile In a trapezoidal profile the motion accelerates linearly up to a maximum speed and then decelerates linearly until it stops A parabolic profile has a parabolic accelerate phase and a parabolic decelerate phase It has the ability to compute several different types of paths like joint interpolated point to point straight line continuous path and relative or blended motion The operating system and control parameters are stored in a non volatile FLASH EPROM 7 memory The robot controller can communicate with other peripherals via two serial RS232 ports 8 with a standard speed of 38400 bps It is possible to expand the controller with additional parallel RISC based processors to provide high speed communication This option was used in the tests in this thesis to connect the robot controller to the real time force and torque sensor equipment see Section 4 In a similar way a vision system can optionally be connected Furthermore for the use of sensors and additional actuators the controller has 16 opto isolated inputs and 12 opto isolated outputs 4 contact relays output and one PLC 9 interface There are four ways of managing the robot controller It may receive commands from the teach pendant the front panel on the controller the terminal emulator on the host computer or from a program executed within the controller
46. s continuous during the sampling This movement 42 is probably closer to reality when activating the buttons but the drawback was that the x location and angle were not measured in exactly predetermined uniform increments The solution was to interpolate the result and thus get uniform x location and angle steps 43 9 References 1 4465 Robot Arm User Manual UMI 17 501 S CRS Robotics Corporation Burlington Ontario Canada 1996 2 John J Craig Introduction to robotics ISBN 0201095289 Addison Wesley Publishing Company Palo Alto California USA 1989 3 Gunnar S Bolmsj Industriell robotteknik ISBN 9144285124 Studentlitteratur Lund Sweden 1992 4 C500 Controller User Manual UMI 17 556 S CRS Robotics Corporation Burlington Ontario Canada 1996 5 http cs stanford edu people eroberts courses soco projects risc whatis index html 6 Torkel Glad Lennart Ljung Reglerteknik ISBN 9789144022758 Studentlitteratur Lund Sweden 2006 7 http en wikipedia org wiki Flash memory 8 http en wikipedia org wiki Serial_port 9 http en wikipedia org wiki Programmable logic controller 10 http en wikipedia org wiki Random access memory 11 RAPL 2 Language Reference Guide CRS Robotics Corporation Burlington Ontario Canada 1996 12 http en wikipedia org wiki BASIC 13 http en wikipedia org wiki Pascal_ programming language 14 Application Development Guide CRS Robotics Corpo
47. solution of the measured force and torque components will be 1 4096 27 4096 When the sensor system measures all six components of force and torque which is the case in these tests the sampling rate in the F T controller is 750 Hz However this rate can be increased up to approximately 3000 Hz depending on the number of components that is measured and the data type used The factor determining the actual update rate in the robot controller however is the serial communication speed rather than the sampling rate In these tests the data is transmitted in ASCII 22 code with a rate of 38400 bps and this results in an update rate of 87 Hz in the robot controller By using binary code instead it 1s possible to increase this to more than 500 Hz The data from the sensor system is made available to the robot controller as part of an array of RAPL interface variables This means that the data is accessible in the same way as any other variables that are handled by the robot controller The reserved template name is FORCE and is indexed from 0 to 61 Some of the elements in the array are read only variables and these consist of the six Cartesian components of force and torque that the transducer is subjected to Among these variables are also the resultant force and torque as well as the maximum sensed forces moments and resultants that has been recorded since the activation of the sensor system All of these mentioned data are well suited for
48. t u b InStr aa DDETal 0 DDETalxx Mid DDETal aa u b aa C InStr 1 DDETalxx 0 Mid DDETalxx C 1 KonvDDELocArrayTillDbl CDbl DDETalxx Exit Function 121 MsgBox Ett fel har uppstatt Prova igen 121 VisaHuvudMeny End Function Function KonvDDEVariableTillHeltal DDETal Konverterar variabel fr n str ng till heltal 49 On Error GoTo 130 A InStr 1 DDETal 0 u b Len DDETal DDETalxx Right DDETal u b A C InStr 1 DDETalxx 9 Mid DDETalxx C 1 KonvDDEVariableTillHeltal CInt DDETalxx Exit Function 130 MsgBox Ett fel har uppstatt Prova igen VisaHuvudMeny End Function Function HamtaDDETal Application Topic Item Funktionen ppnar en DDE f rbindelse med programmet Application hamtar en variabel Item och st nger d refter DDE f rbindelsen Application Tex ROBCOMM Topic Tex VARIABLE Item Tex BUSY On Error GoTo 140 DDEKanal DDEInitiera Application Topic H mtaDDETal DDEBeg r DDEKanal Item DDEAvsluta DDEKanal DDEKommando Application REREAD ALL Exit Function 140 MsgBox Ett fel har uppstatt Prova igen VisaHuvudMeny End Function Sub DDEKommando Application Topic Item Subrutinen ppnar en DDE f rbindelse med programmet Application utf r ett kommando Topic eventuellt pa programnamnet Item och st nger d refter DDE f rbindelsen Application Tex RO
49. the VBA software The movement of the buttons or switches during the sampling of position and force or torgue data could not be done in predefined steps Instead the movement was continuous during the sampling This results in that the x location and angle were not measured at exactly predetermined uniform increments The x location and angle is therefore interpolated into predetermined steps 29 When sampling the results from consecutive tests the mean values together with the maximum and minimum values are stored The robot program Appendix B also consists of several sub modules for initialization testing and recovery The VBA modules activate these programs The predetermined positions for measuring different tests objects are stored in the robot system by teaching the system see Chapter 3 2 30 7 Test Results Analysis and Conclusions The test results are presented in graphical form below starting with the push buttons and then the rotary switches 7 1 Test Results Push Buttons In the following diagrams the distance mm and the correlating measured force N are depicted on the x and the y axes The Swedish terms Vag Kraft Diagram V g and Kraft can be translated to Distance Force Diagram Distance and Force Vag Kraft Diagram Kraft N mm Figure 7 a Push Button 2 I measurement Energy 7 98 mJ Vag Kraft Diagram Medel Min Max K
50. the task of this work The other elements of the array are read or write variables The majority of these variables are used in conjunction with force control mode which is not used in this work 22 S Supervisory Computer System The supervisory computer system is essentially a program written in Visual Basic for Applications VBA 23 which is executed on a PC The system supervises the entire test sequence It manage the user interface commands the robot system to initiate tests then collects and stores the sampled data and finally performs analysis on the data and presents the result to the user The computer is briefly described in Section 5 1 and the software that was used to develop the computer program is presented in Section 5 2 5 1 Hardware The hardware is in this case the same PC that is utilized as host computer by the robot system see Chapter 3 4 which means that the supervisory program and the robot communication program Robcomm run on the same computer This is convenient since it simplifies the necessary interaction between the two programs by enabling them to communicate through the Dynamic Data Exchange DDE interface 5 2 Software Platform The program was written in Visual Basic for Applications VBA version 5 0 which is included in Microsoft Excel version 95 Swedish edition VBA is basically an object oriented language that is found in a couple of Microsoft utility products like Excel as well as i
51. ticulated robot system equipped with a force and torque sensor The test objects have unique force or torque profiles e The tests are repeatable Test objects can be measured and compared to standard objects which have a desired feeling using the demonstrated profiles correlations and inherent energy The standardized and flexible test method has the potential to replace current subjective product tests and improve the development and testing of user interfaces in cars 41 8 Discussion 8 1 Future Perspectives The next step would be to have test persons define desirable profiles for different types of buttons and switches to set model profiles These could then be used for comparison with future test objects The entire system can be developed towards more standardized test procedures and ease of use e g Enhancing the graphical user interface Developing the software for more types of test objects and options Incorporating automatic correlation with model profiles e Standardized fixtures for holding the test objects and applying the test probes at predefined locations Adding entities to better be able to correlate to model objects like maximum and minimum allowed forces maximum slope etc Faster tests by using high speed parallel ports for communication and by using high level programming PCP in the robot programs The standard deviation of the result could be calculated during the sampling
52. ual Basic A programming language from Microsoft Programs designed in VB can be used as stand alone applications Visual Basic for Applications Similar to VB and is built into most Microsoft Office applications It can normally only be used within a host application Table of Contents Abstract Preface Acknowledgments Abbreviations and Terms Table of Contents 1 Introduction 1 1 Research and Development at Volvo 1 2 Problem Definition and Objectives 1 3 Flexible Systems for Product Testing 1 4 Overview of the Thesis 2 Problem Approach 2 1 Measurement Situation 2 2 Desired Measurement Procedure 3 TheRobotSystem 3 1 The Robot Arm 3 2 The Robot Controller 3 3 System Software 3 4 Host Computer 4 TheReal Time Force and Torque Sensor System 41 The Transducer 42 TheF T Controller 5 Supervisory Computer System 51 Hardware 5 2 Software Platform 6 Test Implementation 6 1 Test Objects 6 2 Hardware 6 3 Software 7 Test Results Analysis and Conclusions 7 1 Test Results Push Buttons 72 Analysis Push Buttons 7 3 Test Results Rotary Switches 74 Analysis Rotary Switches 75 Conclusions 8 Discussion 8 1 Future Perspectives 8 2 Specific Challenges 9 References Appendix A Visual Basic Application VBA Code Appendix B Robot Code RAPL 2 Code Appendix C Formulas and Conversions 1 Introduction 1 1 Research and Development at Volvo Volvo the largest car manufacturer in
53. ut two natural properties of feeling are the force and torgue reguired to operate the control The objective of this master thesis was to develop and evaluate a flexible objective and standardized method to measure force and torgue during product testing using an articulated robot system eguipped with a force and torgue sensor A couple of ordinary electrical push buttons and rotary switches were selected to represent the various types of controls A test method comprising the robot system software with graphical user interface and various control systems were developed Several tests were performed on the selected test objects The results demonstrated that it was possible to measure the force and the torgue reguired to operate the push buttons and rotary switches and that they displayed unigue profiles It was also possible to measure the reguired energy to operate the controls The results were repeatable over time and showed little variance From the performed tests and result analysis it was concluded that it is possible to measure the test objects as desired with a flexible objective and standardized method using the articulated robot system equipped with a force and torque sensor Test objects can be measured and compared to standard objects which have a desired feeling and the test method has the potential to replace current subjective product tests and improve the development and testing of user interfaces in cars Preface Persp
Download Pdf Manuals
Related Search