CHAPTER 3 ROBOTIC SENSORS 3.1 DEFINITION AND CONCEPT. * Therefore, some of the important steps to be done by a ‘A robo sensor is a device that detects information about the robot and its surroundings, and at the same time transmits it to the robot's cantraller, Human sensing (Le: eyes to see, hands to physically feet, #215 to hear voice), is considered a¢ the best sensing and Man-made sensors are stil for inferior to human and fother natural sensors in many aspect. lin order to function effectively, a robot has to receive Information from the environment for necessary manipulation, send signals to various joints for necessary Moverient and interact with peripheral equipment. eg. When a robot picks up an ebject and places in a deBinae location, it has to initisily get information about the Presence of the object. As soon as it understands that the object is present, the arm approaches it with a controlled speed and acceleration. While approaching. it ‘Must avold collision with any other obstacle. * NR may also attempt to find the shape and orlentation of the object to be grasped. When the robot grips the object, it must identity the points where It should grip the object with specified force. The object should not be Pressed hard or deformed. oF slip. Sometimes it is ecessary to have pror knowledge about the shape of the object before it is gripped. Therefore, it is required to sense and measure all the important geometrical parameters of the object tying in an environment. Sensory Feedback is thus more important for unstructured enviranmant. el Conerol signa = meet {Nant ren tno a ‘robot in a pick and place operation for which various ‘Sencors, internal as well as external are incorporated = 3 Searching 2 Recognizing > Grasping and > Placing the To Perform Varlous Tsiks without Constant Human ‘Supervision: ‘The human operator it. capable of combining various Types of sensory feedback to perform the most complex Of tasks If industrial robots are to be capable of Performing the same types of tasks without constant human supervision, they too must be equipped with sensory feedback. + Achieve a Greater Degree of Intelligence In Dealing with thelr Environment: ‘The feedback from various sensor is analysed via a ‘Gita! computer and the associated software. The use of ‘Sensing technology to endow machines with greater egree of intetigence in dealing with their environment is an active area to concentrate Therefore, for a robotic ‘manipulator to operate effectively and inteligently and to ‘enable it to work in unstructured environment, it must be ‘equipped with sensory which give information about Inset and ts erwironment. * Intelligent Sensors to Perform Complex Tasks: ‘A robot without intelligent sensors is handicapped and (can only do very specific tasks, while a robot that is able To "see" and “feet” is easier to train to perform complex tasks, * To Provide Positional and Velocity Information of Elements of Manipulator with the Help of Internal ‘Sensors: ‘One of the reason for using sensory feedback in robots i ‘to provide positional and velocity information concer Joint, arm and end-effector status, pasition, velocity and acceleration. This type of feedback is provided continuously and becomes an integral part of the physical robot control system. This category is called ‘intemal © scanned with OKEN ScannerRONOTICS (BE MECHANICAL) on * To Prevent Damage to the Robot Itself, its ee Another reason for using sensory feedback is to prevent paces Cade ap ianerenies Operators and to provide identification and real time information indicating the presence of different types of oer beloeceran yer performed. These involve various sensory devices to suit the needs of the particular robot task being carried out and the characteristics of the working environment This Category of sensors is referred as “external sensors’. ‘The following are the desirable features of vanous sensors © Accuracy: There should not be any error in the Perr petal died geet ne ee ‘@ «Precision: It means, there should net be random variability in the measurement. 0 EET oie cll aoe eee ee © Speed of Responsa: It means that it should be capable of pascioheraproniesntinracgopesny eee pagina en Rallability: It means it should possess a high reliability ee © Cost: It is the overall cost including costs to purchase, perapirtssaneardsaredcomelaes ee ean eae ee ‘The various characteristics discussed below determine the performance, economy, ease of application and applicability (Of the sensor. Therefore, they may be considered before a ‘Sensor is selected. Cost of Sensor: i is an important consideration, for machine using many sensors There should be well balance between the cost and the ather factors ‘= Size of Sensor: Size may be of primary concer Example, in case of joint displacemient sensors are used in design of joints and move with the robot's body elements. The space available around the joint may be limited. Aiso large sensor may limit joint ranges (© Weight of Sensor: A heavy sensor puts an inertia of the am, also redluces its overall pay load. ‘© Type of Output: The out of sensor may be digital or of analog type. The output may be used directly or t may ROBOTIC SENSORS have to be converted, Example, the output of a potentiometer is analog and that of an encoder is digital. ‘Thus, an appropriate type of output be used. Interfacing: Sensors must be interfaced with other Gevices in system. Example, microprocessors and controllers. This is an important consideration. Resolution: It is the minimum step size within the range of measurement of the sensor. Example, in a digital device with‘ bits, the resolution will be, where, Re = 360". R= 225" Sensitivity: It is the ratio of a change in output In response to a change in input. Highly sensitive sensor shows larger fluctuations in output, due to the fluctuations in input combined with noise. Unearity: 1 gives the relationship between input variations and output variations. The sensor with linear output shows that, the same chang within the range will produce the same change But all devices behave in nan-linear ways. Some devices Can be assumed to be linear within a certain range of their ‘operation Some may be linearized through assumptions, Range: Nis the difference between the smallest and the largest outputs of the sensor can produce or the Gitference between the smallest and largest inputs with which it can operate property. Response Time: It is the time that a sensors output requires to reach a certain percentage of the total change. Generally, itis specified in percentage of total change ie. 80% Also it can be defined as the time required to observe the change in output as a result of a change in inp Example: For a simple Hg-thermometer, a response time is long. For a digital thermometer, a response time is short, that measures temperature based on radiated heat. Frequency Response: The frequency is the range in which the systems ability to respond to the input remains relatively high. The larger the range of the frequency response, the better the ability of the system to respond © scanned with OKEN ScannerROBOTICS (RE MECHANICAL) 10 varying input. Thus, it is necessary to consider the frequency response of a sensor and determine whether the sensor's response is fast enough under all operating conditions. © Rellabitty: 1 is the ratio of the number of times a system ‘operates. properly to the number of times it is tied. Therefore, it is necessary to choose reliable sensors that lost a long time. Accuracy: It is defined as how dose the output of the ‘sensor is to be expected value. For a given input, the Output is expected to be a certain value, the securacy i related to how close the sensor's output is to this value © Repeatability: Repeatability is more important than accuracy in most cases. I! the sensor's output is measured, 42 number of times in response to the same input. the ‘output may be different each time, Repemtability is a measure of how varied tne different ‘Outputs are relative to each other. Repeatability is ererally random and cannot be eas com Selection of a sensor for a particular application depends on the following factors. ©The quantity to be measured or sensed. The Interaction of the sensor to be selected with other components in the system. + Expected service Ife of sensor. Level of eophistication involved in the sentor to be selected for a particular application, Various difficuttles assoclated with the use of sensors. Source of power for the sensor. ‘Cost factor associated with sensor. Environment in which the sensors are 10 be used. Malntainabittty of the sensors to be selected is one of ‘the important factors. Ruggedness: itis also an important factor for selecting a rigid sensor. Example: Rugged seniors are being developed to withstand extremes of temperature, shock and vibration, humidity, corrosion, dust and vanous contaminants, fuids, electromagnetic radiation and other interferences. + Avallability: Sensor should be made easily available for qiven application It is the selection criterion. [3.6 CLASSIFICATION OF SENSORY DEVICES Sensors are used for a variety of functions in robotic systems. ‘The most common and minimal use of sensors is to provide information about the status of links and joints of the manipulator and about the working environment of the robot. In addition, sensors can help the robot (for detecting positions and orientation of pars. em ROBOTIC SENSORS @ ensure consistent product quality G to discover variations of shape and dimensions of pans. ( ta identity unknown abstucies and () tedetermine system malfunctions and to analyze it 3.6.1 Functional Classification of Robotic Sensors The major functions of sensors in robots can be grouped into five basic categories which is the functional classification of robotic sensors, Status sensors Environmental sensors Quality control sensors Safety sensors and Workcell control sensors. Fig. 2.2: Functional Clarification of Sensors Status Sensors: The primary aim of sensors on a robot is to sense Positon, velocity, acceleration and or torque or force at (ach joint of the manipulator for position and control ef, motion. [AS these sensors form an essential part of the basic or intemal closed-loop control systems, they are called as “intemal sensors’ or ‘status sensors’. This is because the Internal sensors give feedback on the status of the manipulator itset ‘The effectiveness and degree of accuracy that can be achieved by a manipulator depends on = d= the resolution and 2 the accuracy of intemal sensors Imemal sensors must also be cost effective because they are required for each axis of the manipulator. Environmental Sensors: Sensors are also used to extract features of the objects in the workcell or the surrounding erwironment af the robat. ‘This data is utilized by the computer controller to madify ta a given situation. eg. if the robot has to process several types of different parts, each requiring a different sequence of actions by the robot, such as adjusting the © scanned with OKEN ScannerROROTICS (BE MECHANICAL) ‘gripper orientation or applying the exact gripping force, it must determine the required parameters for each part ‘These types of sensors are placed in the environment of the robot or are extemal to the manipulator and are called ‘externat or “environmental sensors. Following are Some of the Uses of Environmental Sensors: * To detect the presence of workpiece. * To determine the position and/or orientation of workpiece and objects to be articulated, or other objects resent in the workcell © To identity workpiec To determine workpiece properties such 25 size, shape and soon. To detect, identity obstacles in the envionment, and Provide the related information about their site, shape, location, speed and so on. * To provide information about the manipulator environment interaction forces and torques © To provide information about environmental variables such as temperature, humidity and s0 on. © To determine the position and orientation of the end: aetfector, joints, and links of the manipulator. ‘= The information from the extemal sensors would have to be processed by the computer in real-time to guide the manipulator in the execution of its programmed work ole. = Vision system is an important sensing method employed to determine the characteristics 25 — © the pan location. Gi) ofentation of pant Gil) size and shape ete. 32. Quality Control Sensors: = Sensors are employed for some af the important functions as inspection and quality contro The use of sensors permits 100% inspection because sensors can be used to determine a variety of part quality characteristics. The extemal sensors used for environmental feedback can also be used for detecting fauks and faitures in the finished product. = The inspection process can be made a part of the Programmed work cycle and the sequence of operations performed by the manipulator may be linked to the resuit of inspection process. Following Sensors can be Used for Inspection: Computer vision system that can provide a larger variety of information about the workpiece in addition to its Position and orientation. on ROBOTIC SENSORS © Utrasonic sensors. Sonic and other sensors, 4. Salety Sensors: Satery and hazard monitoring is an important function of the sensors in robotics. # The safety of the workers and other equipment in the work environment, and that of the manipulator itself important aspecs, €.g, safety in automation plant If there is power failure, all the links of the manipulator may fall to zero-gravity Position instantaneously, This may injure the human beings in the zone or may damage the equipment or the ‘manipulator itsel A solution is to sense the power failure and apply breaks to prevent the uncontrolled falling af links due to gravity. 5. Workcell Control Sensors: Implementation of interlocks in workcell is another important use of sensors in robotics. + An interiock in the work-cycle is a situation that requires sequencing of tasks such that completion of 2 task must be ensured before proceeding to the next task 9. 2 part must arrive on the conveyor before the gripper of the manipulator can pick it. By incorporating the signals from 2 variety of sensors in the rabot program, the venfication of interlocked task can be done. * Sensors may also be used to detect and resolve the interlocks in the workcell All above categories require sensors to be an integral part of robot control system to accomplish specific control functions. 3.6.2 Sensory Devices Fig. 3.3: Types of Sensors Based on Werking Principle © scanned with OKEN ScannerROOTICS (BE MECHANICAL) as) ROBOTIC SENSORS 3.6.3 Various Sensing Devices used in Robot Workcell’ | In following section, various sensors used in Robot workcel “A robotic workcell may be defined as a Guster of one or | are discussed. ‘more robots and several machine tools or transfer fines that | Sensors Used In Robot Workcells: are interconnected in such a way that they work together it | renewing types of sensors are generally used In robot unison”. 7 workcell, Saved on the oxpan thee art seo man cages of sercar | 3° : 0) Discrete sensors and i) Complex sensors * Eddy Current Detectors: It emits an attemating 1) Discrete Sensors (Simple Sensors): magnetic field at the tip of a proble, which induces © These are also known as simple or digital sensors whose eddy currents in any conductive object in the range ‘output has only two states - orvof, yes/no, high/low, oF 1 of the device. It can be used to indicate presence or (oF Din which case itis frequently referred to as a Binary absence of a eonductive object. Device a rca onnese WaefScs fleas gencess-n hoi © Infrared Sensor: Mt acts as a transducer that tranensits signals from 3 deglatahane vibe inc Ages measures temperatures by the intrared light emitted () Complex Sensors from the surface of an object. It can be used to * The output from complex sensors has more than two indicate presence or absence of a hot abject. state and can be of any value within the range values it * Optical Pyrometer: It is used to measure high measures, temperatures by sensing the brightness of an object's * Complex sensors would include analog sensors and serial surface It can be used to indicate presence or Coeecriorneieen SeNeOrE. absence of a hot object. = Complex sensors interface is the process that tansmits. a mpet sesaetgma tea ee Photometric Sensors: It is used to sense light. It them to digital signals or otherwise conditioning them. lecudes photocell, photoelectric transducers, * Above two categories of sensors used in interface with phototube, photodiodes, phototransistors, and most manufacturing systems can be grouped in three photoconductors. t can be used to indicate presence categories = or absence of an object. (2) Contac sensors © Radiation Pyrometer: It is used to measure high (| More-coeract sensors efi! temperatures. by sensing the thermal radiation (© Process control sensors Fig. 2.4 shows the different types off sensors used in robot workcells femiting from the surtace of an object. It can be used 1 indicate presence or absence or a hot abject. * Vacuum Switches: It is used 10 indicate negative alr nencoese | [Opes wres,| | eocescre | | rors, pressures. It can be used with a vacuum gripper to betewtcion | | "Sommer | | indicate presence or absence of an object. 2 Touch Sensor: © Hlectrlcal Contact Switch: It is a device in which an electrical potential is established between two objects, and when the potential becomes zero, this incicates contact between the two objects, It can be Used to indicate presence or absence of a conductive object © Limi Switehe It is an electrical on-off switch actuated by depressing a mechanical lever or button an the Gevice. 1 can be used to measure presence or absence of an object. © Microswiteh: It is a small-electrcal switeh. It can be used to indicate presence or absence of an object. 3. Force Sensor: © Strain Gauge: It is a transducer used to measure force, torque. pressure and other variables. tt can be used to indicate force applied to grasp on abject. © scanned with OKEN ScannerROROTICS (BE MECHANICAL) 4, Miscellaneous Sensors: © Ammeter; It is an electrical meter used to measure electrical current + Unear Variable Differential Transformer: i is an electromechanical used to measure linear oF angular displacement. © Ohmmeter: tis a meter used to measure electrical resistance. © Pleroelectrle Accelerometer: It is a type of sensor used to indicate or measure vibration. Kis an electrical meter used to pressure and other uid pressures © Thermistor: Iris used to measure temperature based on electrical resistance. 4 Thermocuple: It is used to measure temperatures. It Is based on the physical principle that a juncton of two dissimilar metals will emit an emf which can be related to temperature. 5. Vision Sensors: These are advanced seniors used in conjuncton win pattern recognition and other techniques to view and interpret events that are occurring in the rebot workplace. 6. Volee Sensors: These are advanced sensors used to communicate ‘commands oF information orally to the robot. [3.7_DIFFERENT TYPES OF SENSORS. 3.7.1 Photoelectric Sensors = These are of non-contact type and used for positon Sensing and output of which responds to = (Interruption of alight bean by an objector. G_ Reflection of a fight beam back to its source by an abject or part passing infront of the project beam. = Photoelectric sensors are of four types as shown in block diagram below. Prooncre men Comer Fig. 3.6: Light Source and Receiver Type Sensor on ROBOTIC SENSORS ‘© Principle: “As the work moves between the light source and receiver, the light beam is interrupted, resulting in a switch closure’ (or opening). The light source and receiver type consists of a light source and photoreceiver for a direct sean or through a sean photoelectric control system. The photoreceiv etects ight from the source, ampifies it, and switches its output contacts when a change in nominal conditions occurs. A receiver can be energized on light or energized fon dark depending on its intended applications. When the light path between source and receiver is interrupted, an output in the form of & switch closure is obtained. Output contacts of forms A B and C are widely available with ratings of upto 10 amps. AC devices can switch circuit loads such as relays, solenoids, oF small motors directly. * DC devices are usually used for logic-level switching, the Output swrtches are solid-state and switch DC currents on the order of 150 ma. * Scanning arrangements can be used with source or receiver separation distances of over 30 m, although workcell applications rarely call for such a rang 2 Retrofiector Type: ‘Revco actecte Searmer (uch aves Bom np source bn receive) fig. 3.7: Retrofiective Type Sensor In this type of sensor, both light source and photoreceiver are packaged in the same housing, which is called as scanner. The retrorefiective scanner can switch when the light beam from a fixed reflector is interrupted, or from 2 passing workpiece or machine element. This type of scanner has limited distance between target and light source. 2. Diffuse Scanner Type: (Fig. Lt: Diffuse Scanner Type © scanned with OKEN ScannerROBOTICS (BEMECHANICAL) © In this type of sensor arrangement, both light source and photoreceiver are built into the same housing. © Aworkpiece (as 2 object placed on plane reflective surface) reflects the beam to the receiver when it moves into the sensing position, An output control signal results when the object passes Rhis point. Here light beam is instially aligned to reflect from surface to receiver. Object interferes with beam, thereby initiating a control signal. 4. Reflect Scanner Type: Presermcaner oes Shy nrtoce cnconpece Fig. 3.9: Reflect Scanner Type Sensor © In this arrangement, the sensor works tke 2 combination of the retroflector and diffuse scanner type ©The light source and receiver are both located in the Same enclosure, and the beam uses the pars surtace to reflect light back to the receiver. + The light reflected must be a Gefinitive beam because diffused light alone will not be sufficient to actnate the receiver. The angle of the light projected by sensor is adjusted to improve sensitivity and establish a specif stance at detection. ©The light source is the light-emitting diode (LED), Which generates light in the infrared GR) spectrum. | Most of the sensors use silicon photo-detectors, Which are sensitive to IR light and are economical. = Photoelectnic devices should always be mounted so 5 1o avoid oF be protected from dirty enviconments, The devices should be selected with excess gain G appropriate to the environment Exess Gain Bement 10 (Clean ar 19200 Law contamination 20-1000 | Rederaie cortamanaton 1002500 | Highortamnaton +The doser the working distance, the higher the excess gain characteristics of photoelectric device. on ROBOTIC SENSORS 3.7.2 Limit Switches © i 2 category of micro-switches, which are electromechanical devices actuated either by some part or motion of a robot or machine to alter the electrical circuit + Limit switches are the most common linear position sensors used in robots and work cells. + Limit switch is designed to be mechanically actuated when a machine member or object reaches a particular position. +The machine member physically contacts the limit switch actuator and switches the contacts, normally one to four poles * Thelimin switch can be used to control machine operation sequencing, or to provide machine protection functions by shutting down the machine when machine members or workers are at a prohibited location. ‘There are Following Types of Switches: (© Industrial Umit Switeh: These are usually enclosed in cases in order to protect the switch trom dust, water and human abuse. (W Standard Duty-Enciosed Limit Switch: Thess rugged and enclosed in cases as similar to above ase (Ui) Heavy Duty-Limit Switch: These are available as oil» ght corrosion resistant and are most rugged enclosures than above two cases. These switches should undergo 20 million cycles, without failure, under moderately severe industrial usage © NEMAstype of enclosures are commonly used for limit switches These are used in workcells, Types of actuator Primarily selected on the basis of where the limit switch might be located with respect to the controlled machine 18 is important to install the limit switch in a location and postion such that © The moving member will not destroy the limit switch, © scanned with OKEN ScannerROBOTICS (BEMECHARICAL) GM should be easily accessible for maintenance (6) tris protected trom accidential acruztion. (@¥) Chips, moisture, grease, oF oil does not accumulate on the actuator. Trisillustrated in Fig 310. 3.7.3 Range Sensors Iisa non-contact type of sensor category and is used to sense and measure the distance between the objects and sensing device, Along they can be used to locate Workpiece in the robot workcell * They may be located on the end-effector oF wrist, For robot navigation and obstade avoidance, range sensors are generally used. There are following range sensing nechniques: @)Thangutation Technique (Structured Lighting Approach Gil) Time-of tight Range Finders. ) THlangulation Technique: © [tis a simplest metnod of measuring range. I is explained with the help of Fig. 311 Fig 312 © An object shown above is illuminated by 2 marrow beam of light which is smept over the surface. The sweeping motion is in the plane defined by the line from the object to the detector and the line from the detector to the source. © If the detector is focussed on a small portion of the surface then, when the detector sees the fight spot, its distance D to the ifuminated porton of the ROBOTIC SENSORS From geomety of Fig. 3.12, = Fig. 332 Oo 8 This approach gives a point measurement © W the source-detector arrangement is moved in a fixed plane (.e. up and down and sideways om a plane perpendicular to the paper and containing the baseline), then itis possible to obtain a set af points whose directions from the detector are known, These Gistances are easily transformed to the three dimensional co-ordinates by keeping track of the location and orientation of the detector as the object are scanned, (W) ‘Structured-Lighting Approach: * Inthis approach, the light pattern is projected onto a set of objects and using the distortion of the pattem, 1m caeadae the ange surface can be cakulated from the geometry of "it Fig. 2.1. |. Since @ = be the angie of the source with the basaling. 4 B — be the distance of the source from ot the detector. re, D = be the distance of muminated sive surface from the detector. Fig. LL: Mexsurement of Range using ‘Structured Lighting Approach © scanned with OKEN ScannerROROTICS (BE MECHANICAL) os The light pattem generally employed is a sheet of Tight generated through a cyfindrical lens or a narrow sli Fig, 3.13. shows measurement of range by structured lighting approach. The intersection of the sheet of light with objects in the workspace yields a light strip which is viewed through a television camera. It is placed at a distance of B from the light source. The strip pattern is easily analyzed by a computer to obtain necessary range information. €g. an inflection indicates a change of surface, and a break corresponds to a gap between surtaces. Fig. 3.11 shows a top view of a set of objects and a line from light source which is a sheet of ight In this, arrangement, the light source and camera are placed at the same height and the sheet of ght is Perpendicular to the line joining the origin of the light sheet and the center of the camera lens The vertical plane containing this line is called as the ‘reference plane’, which is perpendicular to the sheet of light. Any vertical flat surface that intersects the sheet of light will produce a verical stnp of light in which ‘every point will have the same perpendicular distance to the reference plane. The basic objective of the arrangement is to position the camera so that every such vertical sinp also appears. vertical in the image plane. Thus évery point along the same column in the image wil be known to hhiwe the same distance to the reference plane. Calibration: It consists of measuring the distance "B between the light source and the lens centre, and then determining the angles Ge and ag. One of the advantages of this technique is that it results in a relatively simple range measuring technique. Once the calibration is completed, the distance associated with every colurnn in the image is computed. (ll) Time-of-Flight Range Finders: Based on time-of-flight concept. there are three methods of determining the range, as under ~ (2) Apulsed-laser system: (©) Acontinuous-beam laser system; and (0) An ultrasonic range finder. ‘Above techniques are illustrated in following section. (3) A Pulsed-Laser Systeme To determine range using a laser, és to measure the time it takes on emitted pulse of light to retum ROBOTIC SENSORS coanially from a reflecting surface. The distance to the reflecting surface is calculated using the relationship eT cs where, Tis the pulse transit time. is the speed of light, A pulsedslaser technique, produces a 2-dimensional array with values proportional to distance, and @ 2-D scan is accomplished by deflecting the laser light via. 2 rotating mirror. So.a 3D object is viewed and it's 20 scan (Le. the corresponding sensed array displayed as an image) is achieved. In 2D sean the intensity at each POM i propomional to the distance between the sensor and the reflecting surface at that point, Darker point is closer, and the bright areas around the object boundaries represent discontinuity in the range etermined by post processing in a computer. The working range of this device is of the order of 1 10-4 m, with an accuracy of # 0.25 em. (b) A Continvous-Beam Laser System: In this approach, a continuous-beam laser is used to measure the delay (Le. phate-shift) between the ceutgoing and returning beams. This technique is as itustrated in Fig. 314 Fig. 1.14: Range Measurement by Phase-shift using Continuous Later Beam Consider a bearn of laser light of wavelength 2 is split into two beams. One of these beams called as ‘relerence bear, travels a distance L to a phas measuring device, and the other travels a distance D out to a reflecting surface Therefore, the total stance travelled by the reflected bear is, Dy = L+20 D = under this condition Dr =k and both the reference and reflected beams arrive simultaneously at the phase measuring device. As D increases, the reflected beam travels a longer path and thus. a phase-shift is introduced between where, © scanned with OKEN ScannerROOTICS (BE MECHANICAL) the two beams at the point of measurement as shown in Fig. 3.14. Therefore, total distance travelled by the reflected beam ig py = te(Qa)a We = 360%, the two waveforms are aigned and me cannot differentiate between [Oy = land (Dr=tent} where n = 1 2, based on measurements of phase shit alone Thus a unique soltion can be achieved only when 0 « 360° 020 «h Since, Oy = Le2D 8 where, 20 355 8 2 ~ 9-(3)8 This equation gives the distance in terms of phase- shift if the wavelength is known, ‘= Limitation of this case that the wavelength of laser Hight Fs small (oF order 632.80 nm for a helaum-ngon lagen, therefore, the method becomes impractical for robotic applications, © Solution to this problem is to modulate the amplitude of the laser light by using a wavetorm of much higher wavelength. In this. case, the reference signal is the modulating function. The modulated laser signal is sent out to the target and the retuming beam is stipped of the modulating signal which is then compared against the reference to determing phase-shift, = An advantage of the continuous beam taser technique vs. the pulse-light technique i that the former yields intensity as well 2s range information. (01 An Uttrasonic Range Finder: ‘= Inis another majar type of the time-of-flight concept = Basic principle is same as in previous case, an Ultrasonic chip is transmitted over a short time period and since the speed of sound is known for a specified medium, a simple calculation involving the time interval between the outgoing pulse and the return echo yielis an estimate of the distance to the refiecting surface. © They are primanty used for navigation and obstacte avoidance, as it introduces severe limitations in resolution due to beam pattem which is around 30°. 2.7.4 Proximity Sensors ‘Most of the proximity sensors do not require any physical contact at all in order to produce a signal that can be 19 0 ROBOTIC SENSORS Used by a robot to determine whether itis near or obstade. ‘The non-contact type of proximity devices depend on variety of operating principles in order to make the Proximity determination. e.g. inductance, magneticeetfect, capacitance ubrasound, optical category etc. These categories are described in the following section, Thus, proximity sensors generally have a binary output which indicates the presence of an object within a object spectied dtance inten Chasiication: = aca] [rasta] [conic] [oman] [Sea SS Sesleoles (‘V2\e(.) ay sty | Peet | St? | | onacag Seer, | | alee | |werencce| | by on seer ||"Escet| | aes atoee sso || sacs || Roam Sore | |e || Practical uses of a proximity sensor in robotics would be = (To detect the presence or absence of a workpart or ‘other object. (2) To sense human beings in the robot workcell. Inductive Sensors: © Principle: Sensors based on a change of inductance Cue to the presence of a metallic object Fig. 3.15 illustrates the operation of inductive sensor. ‘Geure Conscr emmy NN TELA Wi Fig. £15: Inductive Type of Proximity Sensor © scanned with OKEN ScannerRONOTICS (BE MECHANICAL) «An inductive sensor basically consists of a wound ceil located next to a permanent magnet packaged in a simple rugged housing. © AS sensor is brought in close proximity to a ferromagnetic material, it causes a change in the Position of the flux ines of the permanent magnet as shown in Fig. 315 (0) and (0. ‘© For static conditions there is no movement of the fs: lines and thus, no current is induced in the coil © But as a ferromagnetic material enters or leaves the magnetic Rl of the magnet, the resuiting change in the flux lines induces a current puke whose amplitude and shape are proportional to the rate of change of the flux. © The Voltage Waveform: wv High ete votace ‘come ov Fig. 3.16: Inductive Response as a Function of Speed ©The vonage waveform at the output of the col Provides an effective means for proximity sensing. Fig. 3.16 shows how the voltage measured across the Coil varies as a function of the speed at which a ferromagnetic material is introduced in the field of the magnet. ‘+The polarity of the voltage out of the sensor depends ‘on whether the object is entering or lezving the field. ‘Voltage Amplitude V/s Sensor-Object Distance: =. wemoe em ROBOTIC SENSORS noted that the sensitivity falls off rapidly with increasing distance, and that the sensor is effective only for fractions of a millimeter. ‘The sensor requires motion to produce an output waveform, one approach for generating a binary Signal is to integrate this waveform. The binary output remains low as along as the integral value remains below a specified threshold, and then ‘Switches to high when the threshold is exceeded. ) HalL-Effect Sensors: HallEffect: 1 relates the voltage between two points ina conducting or semi-conducting material to a ‘magnetic field across the material, Hall-eflecr sensors can detect magnetized objects only, However, when used in conjunction with a permanent magnet in 2 configuration as shown in Fig. 318 (a), they are capable of detecting all ferromagnetic materials and a Halleeffect sensors arrangement senses a strong magnetic field in the absence of a ferromagnetic metal in the near fet. As shown in Fig. 348 (bl, when a ferromagnetic ‘material is brought in close proximity with the device, the magnetic field weakens at the sensor due to bending of the fed lines through the materia = (2)_Writhout ferromagnetic material Weak magni No J qa {) In presence of ferromagnetic abject "Ir=aeeem + Conventional tee Fig. 317 The relationship between voltage amplitude and Sensorabject distance is as shown in Fig. 217. Its bo Ae © {e) Hal voltage generation Fig. 28 © scanned with OKEN ScannerRONOTICS (BE MECHANICAL) These sensors are based on the principle of Lorentr force which acts on a charged paride trnlling through a magnetic field. The force acts on an axis Perpendicular to the pline established by the direction of the motion of the charged partide and the direction of the field. Lorentz force is given by, F = qtvx8) where,g is the charge, v isthe welocty wector, magnetic field vector, Xs the vector cross product © Mlustratione A current flows through a doped n-type Semiconductor which is immersed in a magnetic field as shown in Fig. 3.18 (C1 As electrons are the majority carriees in retype materials, and the commenbonal Current flows opposite to electron current. The force acting on moving negatively charged particles have the direction as shown in Fig. 318 (€}. This force Would act on the elecvons, which would tend to collect at the bottom of the matenal and thus produce a voltage across it, which is positive at the top. If we bring a ferromagnetic material close to the semiconductor-magnet device, the strength of the material Feld decreases, thus Lorentz forte gets reduced and also the wohage across the semiconductor. ‘+The drop in voltage is the key for sensing proximity with Hall-etfect sensors. the il) Capacttive Sensors: © Prinelple: These sensors are based on detecting a change in capacitance induced by a surface that is brought near the sensing element. = These sensors are capable af detecting all solid and liquid materials, unlike inductive and Hall-effect sensors. Constructional Features Fig. 319 shows basi Components of a capacitive sensor, 36 Fig. 2.2%: A Capacitive Proximity Sersor (2) The Sensing Element (0) The Dielectric Material (0. The Dy Air-Gap (@) Electronic Grauit un ROBOTIC SENSORS (a) The Sensing Element: 1 is basically @ capacitor composed of @ sensitive electrode and a reference electrode. A sensitive electrode in the form of a metalic dist and a reference electrode as aring. (b) The Dielectric Matertat: A metallic disk (Le. a sensitive electrode) and a ring (Le. 3 telerence electrode) are separated by a dielectric material medium. (©) The Dry-Ale Gap: capacitive element to provide isolation. The Beetronic Cireuit: The sensor is composed of electronic circuitry which can be included as an integral part of the Unit, wherein it is normally embedded in a resin to provide seating and mechanical support Operation: There are varieties of electronic approaches for detecting prosimity based on a change in capacitance: 3 First approach includes the capacitor as part of an Coscilator circuit designed so that the oscillation starts only when the capacitance of the sensor exceeds a predefined threshold value. The stant of oscillation is then tvanslated into an output voltage which indicates the presence of an object. This method gives a binary output whose triggering sensitivity epee nthe theshold vale. 6 ayn eve | 2 —__ 5 15 ewes Fer prod nace (nm) Fig. 320: % Change in Capacitance ws. Distance 2 The second approach utilizes the capacitive element 2 part of a cireuit which is continuously driven by a reference sinusoidal waveform. A change in © scanned with OKEN ScannerROOTICS (BE MECHANICAL) capacitance gives a phase shift between the reference Signal and a signal Gerived from the capacitive element [The phase shift] « [The change in capacitance] Thus, it can be used as a basic mechanism for proximity detection. Fig. 3.20 shows how capacitance varies a¢ a function of distance for a proximity sensor based on the above concepts (ie. approaches). The sensitivity decreases sharply past 3 few millimeters, and that the slope of the response curve depends on the material being sensed. = These sensors are operated in a binary mode so that a change in the capacitance greater than a preset threshold indicates the presence of an object. while changes below the threshold indicate the absence of an object with respect to detection limits established by the value of preset threshold, v) Ultrasonle Sensors: © Principle: “it operates on the principle of reflected Ultrasound, wherein, the time between the signal output and retumn of the reflected signal is proportional to distance between the sensor and the target’ * Structure of 2 Uttratonle Sensor: A typical utrasonic sensor used for proximity sensing consists of the following components (2) An Blectroacoustic ranscucer, () Resin layer, (@) Acoustic absorber, and (@) Housing 2 Sensor housing Fig. 221: An Unrasons Promimity Sensor (a) An Electroacoustic Transducer: Ils a basic element of an ultrasonic proximity sensor and is of ‘the piezoelectric - Ceramic Type: It is used for transmitting and receiving ‘acoustic signals, {b) The Resin Layer: =f protects the transducer against humidity, dust, and ther environmental factors. = Malso acts as an acoustical impedance matcher. eu ROBOTIC SENSORS (©) Acoustic Absorbers: * In order to detect abjects at dose range, fast damping of the acoustic energy is necessary, which is accomplished by prowiding acoustic absorbers. © One of the absorber it mainuined at the ceramic transducer and the other acoustic absorber used around the metallic housing (4) The Housing: # The housing is designed so that It produces 3 narrow acoustic beam for efficient energy transfer and signal- Girectionatity. + tis of two kinds, one is metallic housing which separates acoustic absorbers and other is sensor housing which gives mechanical support to sensor. Operation: * Different waveforms are analyzed, which are used for both transmission and detection of the acoustic energy signais Fig. 322 shows typical set of waveforms used for the Fig. 222: Different Waveforms Related to an Ultrasonic Proximity Seror It is the gating signal used to control © Waveform * transmission. * Waveform 'B': It shows the output signal as well as the resulting echo signal, # Pulse °C: The pulse in °C result ether upon transmission or reception. * Waveform "D': In omer to differentiate between pulses corresponding to outgoing and retuming energy, a time window is introduced, called 2s waveform ‘D’, which essentially establishes the detection capability of the sensor. © scanned with OKEN ScannerRONOTICS (BE MECHANICAL) Le. time interval {31 isthe minimum detection time, and (at, + 41) is the maximum detection time. is noted that these time intervals are equivalent to specifying distances since the propagation velocity of an acoustic wave is known given the transmission medium. # Wavetorm ‘E: A echo received while signal ‘D' is high Produces the signal E which is reset to low at the end of a transmission pulse in signal 'A\. © Wavetorm “F: Signal is set high on the pastive edge Of a pulse in’E" and is reset to low when "Eis tow and a pulse occurs n'a. In this way, will be high whenever an object is present in the distance interval specified by the parameters of waveform 'D’ Thus, Fis the output off interest in an Ultrasonic sensor operating in a binary mode. iv) Optical Proslmity Sensors: © Principle: ‘Optical proximity sensors detect proximity of an object by its influence on a propagating wave as it \ravels from a transmircer to a receiver, 25 similar to that of an ultrasonie sensors” © Optical sensors can be designed using either visible oF Invisible light source. Invisible (Le. infrared) sensors may be Active of Passive The Active Sensors send out an infrared beam and respond to the reflection of the beam against a target This kind of sensor can be used to indicate not only whether or nor a partis present. but also the postion of the part By timing the interval from when the signal is sent and the echo is recetved, 2 measurement of the distance between the object and the sensor can be made. This aspect is essentially useful for locomotion and guidance systems, © Passive infrared sensors are simply Gevices which detect the presence of infrared radiation in the environment These sensors often find application in security systems to detect the presence of bodies giving off heat within the Fange of the sensor. Fig, 323 shows one approach for detecting proximity by Optical means aw ROBOTIC SENSORS «This sensor consists of a solid-state light emitting diode (LED), which acts as a transmitter of infrared light and a solid-state photodiode which acts as the receiver + The cones of the light formed by focussing the source and detector on the same plane intersect in a long, pencillike volume, This volume defines the field of operation of the sensor since a reflective surface which intersects the volume is illuminated by the source and simultaneously ‘seen’ by the receiver. © Its important to;nate that the detection volume shawn in Fig, 323, does not yield a point measurement (le, 3 surface located anywhere in the volume will produce @ reading) while it is possible to calibrate the intensity of these readings as a function of distance for known object COnientations and reflective characteristics. 3.7.5 Touch Sensors Touch sensors are used to obtain information associated with the contact between a manipulator hand and objects in works pace. # The touch information can be used for ~ 2 object location and recognition, and 2 19 control the force exerted by @ manipulator on a given object * Touch sensors are categorized into ~ (a) Binary sensors and (©) Analog sensors. (a) Binary Sensors: + Binary sensors are basically switches which respond to the presence or absence of an object = Binary touch sensors are contact sensors such as micro-swtiched. As shown in Fig. 3.24, a switch is placed on the inner surface of each finger of a manipulator hand. Fig. 3.24: Robot Hand with Binary Touch Sensor © scanned with OKEN ScannerRONOTICS (BE MECHANICAL) This type of sensing is used for determining if a part is present between the fingers. It is also possible to center the hand over the object for grasping and ‘manipulation, by moving the hand over an ebject and Sequentialy making contact with its surtace. © In order to provide further tactile information mutiple binary touch sensors can be used on the inner surtace of exch finger. Also they are often mounted on the extemal surtaces of a manipulator hand to provide control signals Useful for guiding the hand thoughout the workspace, fb) Analog Sensors: «An analog sensor is a compliant device whose cutput 15 proponional to a local force, Fig. 325 shows simple arrangement of an analog device consisting of a spring-leaded rod, which is mechanically inked to a rotating shaft in such a way that the displacement of the rod due to a Lateral force ‘esuns in a proportional rotabon of te shah. seryee ee a Fig. 3.25: An Analog Touch Senior = The rotation is then measured continuously using 2 potentiometer or digitally using a code wheel, © Further based on the spring constant, it yields the force corresponding to a given displacement. ‘Tactile Sensing Array: = Tactile sensing amays capable of yielding touch information over a wider area that afforded by a single ‘© Fig. 326 shows use of these devices, with a robot hand in Which the inner surface of each finger hrs been covered With a tactile sensing array. = The extemal sensing plates are binary devices and have the function as discussed in previous section_ 15 Fig. 1.26 & Robot Hand with Tactile Sensing Array [3.8 GLOBAL POSITIONING SYSTEM (GI ‘The Global Positioning System (GPS) is the most common mechanism used for location estimation. It provides a 3D position estimate in absolute co-ordinates as well as current lume and date and ts available anywhere on the Earth's surtace. Standard GPS provides a potition estimate in the horizontal plane to within about 20 m range. t was developed for miltsry applications but has become widely adopted In crvilan applications, including automobile navigation systems, recreational ofienteering, and inventory tracking for transportation companies Introduction: The system és based on received radio signals transmied by _an ensemble of satellites orbiting the Eath. By comparing the time celays from the different satelite signals, a position fix canbe computed. GPS and Satellite Systems: = NAVSTAR satelite ‘The GPS system is based on the NAVSTAR satellite system Geployed and maintained by the United States, spectically by the Air Force Space Command. * GLONASS Satetite Russian government operates similar system named Globainaya Navigatsionnaya Sputrikovaya Sistema (GLONASS) but at the time of writing it is not available for robotic applications. * Galileo Satetine European Union operates a system named Galileo, and is expected to offer two different classes of service: an open service and an encrypted higher-quality commercial service. + Beidou and QZ5S Other GPS. systems such as the Chinese Beidou and the Japanese QUSS are available. © scanned with OKEN ScannerWHOTICS (RE MECHANICAL) In general the term GPS always refers to the NAVSTAR system, NAVSTAR provided two different services The Precise Positioning System (PPS) reserved primarily for milltary users, and the Standard Position System (SPS) with lower accuracy. ‘The GPS Satelite Network his based on a base constellation of 24 orbiting satellites along with up to six supplementary additional satelites that are also operational. These satelites: are in ‘almost-cireular medium Earth orbit. As opposed ta being geostationary the orbits are semisynchronous. ie. their Position relative to a ground observer is constantly changing, and that their orbital period is exactly half a sidereal day. ‘The orbits are selected 0 that from almost any point of the Earth's surface there will sheays be four or more satelites directty visible 0 is referred as a criterion for cbtaining a GPS position estimate) The satellites are ‘organized into six orbital planes with four satelines in ach The system Is designed such that averaged Ove the tntire Eann’s surface and over a 24h internal, the satelites should allow 99.9% coverage at the worst covered location in any 24h interval, and the signal should be at least 83.92% available at the worst place on Earth on the worst day over a 30-day measurement Interval Each satelite repeatedly broacicasts a data packet known aS the Coarse-Acquisition (C/A) code, which is received by the GPS receiver on the Li channel at 1575 MHz. The simple principle is that, f the receiver knows the absohvte positions of the observed satelites, the receiver position an be directly determined. If the signal propagation time for the radio signals were known, the receiver Position could be computed directy via triateration as shown in Fig. 327 below. This implies that an absohite timing reference is present ‘on the receiver, which would be prohibitively costly Instead, only the satellites have highly accurate atomic ‘docks (accurate to approximately 1 in 300 000 years) computes the difference in sgral Propagation times between the diferent satelites, and Uses this to compute a range estimate refered to as a Pseudo-range (to explicitly indicate that it is corrupted by several sources of measurement noise). The receiver au ROBOTIC SENSORS + The specific geometric problem is referred to as ‘multlateration or hyperbolic positioning and the ‘solution is computed using a sophistical ‘Kalman fier? within the GPS receiver. To avoid retaining an ephemeris {0s} table for the satellites and a very accurate clock in the receiver, each satelite broadcasts its own position ‘and an accurate time signal as part of the data packet tacit transmits, OK Fig. B27: Titateration of GPS + GPS trilateration on the plane. Suppose that one receives Signals from three transmitters (A, B, and C) with known locations. Knowdge of the signal delay from one emitter (say A) localizes the receiver to lie on a circle of known diameter (0,) whose center is the emitter, The constraints from two emitters intersect at two points (maximum). A third emitter is required to disambiguate these two solutions. In three dimensions, the signal propagation constraint trom a single emitter constrains the receiver to lhe on a sphere. The intersection of the constraints from wo emitters constrains the receiver to lie on a circle, The intersection of the constraints from three emitters Constraints the receiver to one of two points. * GPS satetines broadcast at several different frequencies known as LU through LS; only L2 (1575.42 MHz) and 12 (1227.6MH2 are used by civillan GPS receivers. The standard service offered by NAVSTAR and the performance criteria for it is determined by the L1 signal, which contains two unencrypted components: The acquisition message (coarse-acquisition message C/A) and a navigation data message. It is also possible to use the encrypted 2 signal as well, even without the secret Gecryption keys, to provide augmented error correction (by observing the relative effects of ionespheric distortion as a function of frequency). The restricted-access signal broadcast on both the Li and L2 channels is the P-code (as weil as a fairly recent M-code) which is known as the Y-code or PIV) or PAY code once it is encrypted. Both the C/A and PY) codes indude the navigation message stream that spectfes clock bias data, orbital information, © scanned with OKEN ScannerRODOTICS (BE MECHANICAL) am Jonospherical propagation corrections factors. ephemeris data, status information on all the satelite, universal time code, and other information. ‘©The satellite performance is coordinated by the master Control station located at the Schviever Air Force Base near Colorado Springs. Colorado, USA and is connected to agicbal network of five additional monitoring stations (Cape Canaveral USA, Ascension Island, Kwajalein Atoll in the Marshall Islands, Diego Garcia Atoll and Hawai) which are used to produce the measurements that are Uplinked to generate the navigation message sveam, Finally, it should be noted that an additional signal is now available on the 12 frequency band. This LIC signal on satellites designated block TR-M promises to provide much improved receiver sensitivity so that position fixes can be abuined in environments, such as in forests, where they are currently not readily available. Performance Factors: GPS performance depends on several factors: (a) Satellite transmission accuracy, (©) Environmental conditions, (0) Interactions with ground-based obstacles, and (a) Receiver properties. In the context of robotic, factors that atect the Performance of the satellites themseves and the atmospheric conditions are essentially uncontrotable, Nevertheless, it should be noted that these can be Sourtes of error and that the GPS signal isetf may not abvays be reliable. A service failure ts defined as a set of Circumstances where the positioning service exhibits 2 typical error behavior (ie. incorrect signals). Such failures are classified 2s either minor or major failures. Minor failures are those that have limited impact on a receiver and lead to ranging errors of under 150m. Major failures are those that lead either to larger errors or data processing avertoads in the receiver. if a single satetite experiences an error that leads to a major failure, then within a6 h period, approximately 63% of the Earth's surface will have the satelite in view at some point © The controllable factors in using GPS for accurate hocalization are (2) It requires an unobstructed fine of sight to the satellites, (©) Itdepends on atmospheric conditions, and (©) It depends on the ability to receive (weak) radiofrequency communications. = There is a potential for wildly incorrect estimates. Generally satellites that are directly overhead provide ROBOTIC SENSORS better signals than those near the horizon. In addition, since the basis of GPS position is differential signal analysis, it is best if the satellites used in the GPS Computation are widely spaced in the sky. GPS signals are in the microwave band and, as such, they can pass through plastic and glass, but are absorbed by water (wood, heavy foliage) and are reflected by many materials AS a consequence, GPS is unreliable In heavy forest, deep canyons, inside automobiles and boats, in heavy snowfall or between tall buildings. In some cases, partial obstruction of the sty may not prevent @ postion estimate from being computed. Assuming the minimum eumber of satellites operating at any time is 24, then on average across the Earth's surface ‘ight satellites are in view 50 that even partial occlusion of the sky can often be tolerated. On the other hand, partial occlusion can lead to reduced accuracy since the selection of available satellites used for computing position becomes limited and optimal accuracy is obtained by Using as many satelines as possible (weighting them appropriately in the intemal Kalman fiter. Secondary factors that cifferentiate different GPS recervers are the rate at which the signals are collected, the receiver sensitivity, the number of satellites used in the final computation, the number of factors taken into account in the estimator, and the exploitation of supplementary positioning schemes such as the Wide- ‘Area Augmentation System (WAAS). [A major factor in determining the cate at which estimates can be produced is the number of independent receiver elements in the GPS system. Sequential singleechannel feceiwers are Simpler and thus more econornical (and potentially smatien, but they must lock sequentially onto each satellite being used. Parallel multichannel receivers an lock onto to more than one satellite at once, and are generally faster and more costly; some degree of parallelism is the norm in good-quality consumer devices. GPS computations are based on an estimation of the Dikrtion of Precision (DOP) and specifically for the dilution of precision of the positional parts of the system, |. @, Positional Ditution of Precision (POOP), These correspond 10 partial derivatives of the error with respect to position and allow the mast accurate ensemble of visible satellites to be determined at any time. The standard implementation for GPS systems specifies that POOP ‘values be recomputed every 5 min. ‘The minimum performance parameters for GPS receivers are based upon transforming instantaneous. range residuals to a user position estimate using a linearized position sation from a stationary, surveyed focation, © scanned with OKEN ScannerRONOTICS (BE MECHANICAL) © Most GPS receivers use additional techniques such as ange residual smoothing, velocity aiding, Kalman filtering, of multiple satellite (allinwiew satelite) Solutions That said, formal performance for the system is ‘measured with respect to the minimum. ‘The GPS Peshion Estimation Algorithm Ris summarized as follows ‘+ Select the best four satellites based upon the minimum feror measured in terms of POOP. Update every five minutes, of whenever 3 satelite being Used in the solution sets. ‘© Measure the pseudorange to each satelite. Each of the four measurements must have a reception time tag within O55 5 of the solution time. The reception time tag is based upon measurement system time, and the transmission (Ume tag is based upon satelite time, Determine the ephemeris for each of the satedites Being used, and compute the Eanh-Centered, Earthficed (ECEF) co- ‘ordinates for each. Correct for the Earth's rotation and thus compute an estimated pseudorange measurement that ‘should be obtained for each satetice Compute the range residuals as the differences betmeen the ‘actual and observed measurements. Estimate the matrix G that determines the overall system solution, known as the pasition solution geometry matrix. The matric can be desenbed in terms of 2 collection of row vectors, ene for each of the satellites being used, each row being made up of the x y, z and time coordinate drecton ‘cosines for the vector between the user and the satelite (with respect 10 a fixed reference frame for the planet called the World Geodetic System, WGSB4). ‘Compute the user's position: ‘The standard implementation of GPS is based upon a pesition fix rate of ence per second, although faster and slower rates ‘are possible. Under typical operating conditions and without ‘Specialized enhancements GPS accuracy is roughly 20-25 m in ‘the horizontal plane and 43 m in the vertical cirection. The restricted PPS signal provides an accuracy of at least 22m {typical values are 7=10 m) in the horizontal plane and 27.7 m in the vertical direction as well 2s coorinated universal time (UTC) time accuracy within 200 ns based on a reference signal {from the US Naval Observatory GPS signals can be affected by mutipath issues where the radio signals reflect off surrounding terrain - buildings, ‘canyon walls, hard ground, etc. This delay in reaching the receiver causes inaccuracy. A variety of receiver techniques, most notably narrow correlator spacing [20.20 have been eveloped to mitigate multipath errors For tong delay multipath, the receiver itself can recognize the wayward signal aun ROBOTIC SENSORS and discard it To address shorter delay multipath due to the signal refiecting off the ground, specialized antennas may be used. This form of mutipath is harder to filter out as it is only Slightly delayed s compared to the direct signal, causing effects that are almost indistinguishable from routine Auctuations in the. del * Wis. a cevice tat uses measurement systems such as gyreseopes and accelerometers 10 estimate the relative position, welocity, and acceleration of a vehicle In motion. The resutting navigation system thus formed is known 3s an inertial navigaton system of INS. It was demonstrated in 1949 by CS. Draper, Now IMUs have become a common navigational component of aircraft and shi DMUs come in two basic categories, 1, Gimbaled systems and 2. Strap-down systems. L_ Gimbaled IMU: These are mounted within complex gimbal structures in order to provide a stable platform from which measurements can be made, Gyroscopes sed to ensure that the gimbal remains aligned with the instal reference frame at power up, The orientation of the gimbaled platiorm relative to the vehicle Is used to map measurements taken within the IMU to the reference frame of the vehicle 2. Strap-Down IMUs: They have the IMU rigidly connected to the vehide (strapped down}, to no such transformation is required. In wither case estimating the motion relative {0 the inital frame requires integrating information from the sensors within the IMU (accelerometers, gyroscopes, fc) in real time. With the low cost of computation today, and the costs associated with manufacturing and Operating gimbaled IMUs, strap-down IMUS are_much more common today. A true IML maintains a 6-Degres- OF-Freedom (DOF) estimate of the pose of the vehicle: asition (ky, 2) and orientation (Le. roll pitch, yaw). ‘Oreraaton wagon ] rare] frees | neon vu ‘ea me weoay postion Rate gyros | I So mens] prem poi, ly les Fig. 1.28: Inertial measurement unit © scanned with OKEN ScannerRODOTICS (RE MECHANICAL) * In addition to maintaining 3 6-DOF pose of the vehicle, commercial IMUs also typically maintain estimates of velocity and acceleration. ee res tea een Fig. 3.28. This IMU uses three orthogonal accelerometers, and three orthogonal gyroscopes. The gyroscope data (es) Cupra ares oars orientation (6). At the same time, three accelerometers: are used to estimate the instantaneous vehicle acceleration (a). This data is then transformed wia the Soo tea eee gravity, s0 that the gravity vector can be estimated and ee ee acceleration is then integrated to obtain vehicle velocity {W) and then integrated again to obtain the position (¢. ate erst came ee the underlying gyroscopes and accelerometers. Drift Ses ee See orientation relative to gravity, resulting in incorrect Soa gee eae ea ieee aions result in a quadratic error in position As it is never possible to eliminate the gravity wector completely, and this any other error is integrated over tme, dnt is a pe ee ae ee ea eee, ‘as he most pawerful of rabot sensory capabilities. Definitions: "Robot wision may be defined as the process. ‘Of extracting, characterizing and interpreting information from images of a Urree-dimensional world”. or dpeanporonpepycsiae ere system must meet the following criteria: > The need for a relatwely low-cost vision system (typically uncer 15 lakhs). > The need for relatively rapid response time needed for robot or manufacturing applications, typically a fraction of a second: ees Simplicity: and Bee ait 3. 1 Need of a Vision in = Robatic ee ee ae include selecting parts that are randomly oriented from a anes © Itisalso used for parts Kientification. 1 eae a ete cae ca Mee ie yyy a1 ROBOTIC SENSORS * Sometimes, tis used in traditional applications to reduce the cost of part and too! fituring, and to allow the robot rogram to test for and adapt 10 limited variations in the evironement. * Advances in vision system, enhance, the vision capabilities, 10 allow for vision-based guidance of robot arm. + Iris used for complex inspection for close dimensional tolerances, improved recognition, and part location capabilities. * Advances in vision systems will permit applications not only in manufacturing, but also in photointerpretation, wave-housing. robotic operations in hazardous environments, autonomous navigation, cartography, and medical image analysis, + Therefore, vision systems can provide information about the position, orientation, identity and condition of each pam in the surroundings. This data ean be used 10 automate the manipulation of objects, plan robot mations to avoid collision with obstactes, or decide how to grasp an object. + Use of robotic vision system makes assembly, quality control parts handling, and classification tasks more robust 3.10.2 Robot Vision System - Levels of Processing Robot vision system has three broad levels of processing as = (Low level vision system: (3) Medium level vision systems; and (a) High-tevel vision system. ‘Above three levels are subdivided inta: (Low Level Vision Systems: Low level vision systems are primitive in the sense that they may be considered as “automatic reactions” requiring no intelligence on the part of the vision system. Low level vision syterns are subdivided into: (a) Sensing: It is the process that yields a visual image. Special lighting techniques are frequently used to obtain an image of sufficient contrast for later processing (b) Preprocessing: It deals with techniques such as noise reduction and enhancement of details. (@ Medium Level Vision System: Medium level vision systems are those processes that extract, ‘characterize, and label components in an image resulting from low-level vision. Medium level vision systems are subdivided intcr (2) Segmentation: It is the process that partitions an image into objects of interest. © scanned with OKEN ScannerROLOTICS (BE MECHANICAL) 20 {b) Description: It deals with the computation of features (eg. size, shape etc) suitable for differentiating one type Bt ciect ore ence (©) Recognition: It is the process that identifies the objects (eg. wrench, bott, engine block). {ll) High Level Vision System: Tk is the system that refers to processes that attempt to ‘emulate cognition Interpretation is weated as high level vision system, Interpretation: Irassigns meaning to an ensemble of recognized objects. 3.10.3 Functions of a Machine Vision Systerns. ‘Machine vision is concemed with the sensing of vision data ‘and its interpretation by a computer. The operation of the vision system consists of three functions: 11. Sensing and digitizing image data. 2. Image processing and analysis. (2) Preprocessing — Low level vision, (0) Segmentation (@) Description = Medium level vision (a) Recognition (©) Interpretation — High level vision 3. Applications: (2) Inspection of pare (©) Identification of part and (©) Visual sensing and navigation. Machine vision fs concemed with the sensing of vision cata and its interpretation by a computer. The typical vision system consists of the camera and digitizing hardware, a digital ‘computer, and hardware and software necessary to interface them. The operation af the vision system consists of three functions as illustrated. 1. Sensing and digitizing image data 2. Image pracessing and analysis 3. Application 2. Sensing and Dightsing tmage Dats: The sensing and digitizing functions invoive the input Of vision data by means of a camera foassed on the Scene of interest. Special lighting techniques are frequently used to obtain an image of sufficient Contrast for later processing. The image viewed by the camera is typically digitized and stored in computer memory. ‘+The digital image is called a frame of vision data and is frequently captured by a hardware device called 2 grabber. = The frames consist of a matrix of data representing Projections of the scene sensed by the camera. The elements of the matrix are called picture elements, or ROBOTIC SENSORS pixels. The number of pixels are determined by 2 sampling process performed on each image frame. A Single pixel is a projection of a small portion of the scene which reduces that portion to a single value, ‘The value is the measure of the light intensity for that lement of the scene. Each pixel intensity is converted ino a digital value, +The digitized image matrix for each frame is stored 43nd then subjected to image processing and analysis functions for data reduction and interpretation of the image. These steps are required in order to permit tealtime application of vision analysis required in robotic applications. * Typically an image frame will be thresholded to produce a binary image, and then various feature measurements will further reduce the data fepresentaton of the image. ‘+ This cata reduction can change the representation of a frame trom several hundred thousand bytes of raw image data to several hundred bytes of feature value data. The resultant feature data can be analyzed in the available time for action by the robot system. Image Processing and Analyse: Various techniques to compute the feature values can be programmed into the computer to obtain feature descriptors of the image which are matched against Previously computed values stored in the computer. These descriptors inchude shape and i characteristics that can be readily calculated from the thresholded image matrix * To accomplish image procesting and analysis, the vision system rmust be trained frequently. In taining, information is obtained on prototype objects and stoned a5 computer models. + The information gathered during training consists of features such as the area of the abject, its perimeter length, major and minor diameters and similar features. During subsequent operation of the system, feature values computed on unknown objects viewed by the camera are compared with the computer models to determine if match has occurred. ‘The final function of a machine vision system is the applications function. The current applications of machine vision in robotics indude inspection, part ‘identification, location and orientation. Research is fengoing in advanced applications of machine vision for use in complex inspection, guidance and navigation Many two-dimensional vision systems can operate on a binary image which is the result of a simple ‘Wwresholding technique. This is based on an assumed © scanned with OKEN ScannerROBOTICS (BE MECHANICAL RoROM SENSORS high contrast between the object(s) and the | 2.10.5 System Hardwares and Funetions Background. Image contrast can be manipulated by Using a controlled lighting systern, Facto | =] Another way of classifying vision systems is according to the number of gray levels used to characterize the image In a binary image the gray level values are PiSeang ort fa) Sorat etry | comanion divided into either of two categories, black or white semete | Sameeg Other systems permit the classification of each pixel’s Canter Gray level into various levels. the range of which is *Excatieg aalled a gray scale. imap * As is true in humans, vision capabilities endow a eae FOBOR With & sophisticated sensing mechanism that ~ allows the machine to respond to its environment in A lpeg an “intelligent? and Rexible manner. a Sruchred 3. Applications of Vislon Include: ~ + Detecting object presence oF type. ste * Determining object location and orientation before ‘aan ite grasping. amen * Feedback during grasping. oe Feedback for path control in welding and other cd continuous proces ses. anne * Feedback for fitting a part during assembly, oo © Reading identity codes, * Object counting, Ld (Wye * Inspection, example of printed circuit boards to rawaee |Pece detect incorrectly inserted components. ee 2.104 A Typical Vision System fora Robot aes A typical vision system for a robot és shown in Fig. 3.29. A CCD pen (Charge-Coupled Device) camera is placed which views the eBeptertato: object. The image is projected by a video camera onto the + Tretening CCD which detects, stores, and read out the accumulated + Ragen ‘charge generated by the light on ach portion of the image. ong Light detection occurs through the absorption of light on + op Photoconductive substrate (¢g. silicon). smcen tafe verter + Despre (ma marin eo iE sees} eroee 7 = = Ee oupee rachrg To wenanses | (Artes |) rapcin Fann (a eartcnon i | (3 Vat esot |__| compe ea en =— Fig. 1.29: Robot vision system-block diagram © scanned with OKEN ScannerEnd Effectors: Robots come in variety off sizes, shapes and capabilities. As ‘discussed earfier robot have four Basie components. © Amanipulator. * Anend effect which is part of the manipulator. © Acomputer controtier and © Apower supply, Robot with above components is as shown in Fig. 21 1g. 2.1: A robot system components ‘The manipulator Is a mechanism that consists of various ‘segments or arms and composed of three sectons: 1, The major linkages. 2. The minor inkages. 3. Theend effector (gripper or tool. The mijor linkages are the set of jointlink pairs that out- Patition the manipulator in space. They consist of the frst ‘three sets, The minor linkages are those joints and links assoclated with the fine positioning of the end-effector. They Provide the ability to orient the tool mounting plate and ‘subsequently the end-effector once the major linkages get it ‘close to the desired position. [2.2 TYPES OF END-EFFECTOR Definition: “A device, which is mounted on the toot plate, and ‘used to make intentional contact with on object oF to produce the robot's final effect on its surroundings by performing a particular task". “Thus, it acts as the bridge between the robot arm and the ‘environment around it. The actions of the gripper vary ‘depending on the task. CHAPTER 2 ROBOT GRIPPERS * Arobot end-effector which is attached to the wrist of the robot arm is a device that enables the general-purpose bot to grip material, pans and tools to perform a ‘specific task. * The end-etfectors are ako called as grippers. End effectors can be broadly divided into two major categories. | oles obo] 221 ‘These are end-elfectors used to grasp and hold objects. Th may be two. oF more fingered devices designed to grasp an eBjact oF tool in a manner similar to the human hand and fagee # The gripper ray be a simple pneumatically controled Geviee that opens and closes or a more complex servo Controlled unit capable of exerting épecified forces or of measuring the part within ts grasp * Grippers may be designed as physical constraints or as fecbon devices ©) A physical constraint device might work like a spatula ‘hat slides under an abject to enable one to lift it. ii) frictional device depends upon the frictional force Between two materials to provide the gripping forc 22.2 Classification of Grippers Grippers can be classified bated on = 1 The number of gapping devices mounted on the robot's writ (a) ‘Single gripper and {}) Double gripper (c) Muhiple gripper 2. The mace of grpping (a) Intemal gripper (®) Extemal gripper 3. The number of Degrees at Freedom (DaF) Incorporated in the gripper smucure (a) 1 DoF (b) 2DoFs LL The Number of Gripping Devices Mounted on the Robot's Wrist: (a) Single Gripper: It is distinguished by the fact that fenly one grasping device is mounted on the rebet's wrist and is used to handle single object. ey © scanned with OKEN ScannerROBOTICS (BE MECHANICAL) en ROBOT GRIPPERS. {b) Double Gripper: It has two gripping devices attached to the wrist and is used to handle two separate Objects. It is useful in machine loading and unloading applications. (©) Multiple Gripper: if two or more grasping devices are fastened to the wrist, then its a multiple gripper, where double grippers are 2 subset of multiple gripper. Multiple gripper systems enable effective simultanecus execution of more than two different jobs. 2. Mede of Gripping: (2) Internal Gripper: This tind of gripper grasps the part fon its internal surface, therefore called) as “intemal gripper. is.as shown in Fig. 22. {b) External Gripper: This king of gripper grasps the part Gn its external surface, therefore called as “external ripper. Ik is as shown in Fig. 23. Fig. 22: Internal gripper Fig. 2.3: External gripper 3. The Number of Degrees of Freedom (DoF): (a) 1 DoF: Some of the mechanical grippers belong To the class of 1 DoF. (b) 2 DoFs: Frew grippers belong to the class of 2 DoFs 4. Special Grippers: ‘The grippers of the robot may be specialined Gewices lke Remote Centre Compliance (RCC) to insert an external mating ‘component into an internal member like inserting a phag into ‘anole. 2.2.3 Process Tooling ‘© Its an end-effector designed to perform work on the part rather than to merely grasp it. ‘© Thus process tooling may be any useful device such as (Asspotewelding torch (i) A spray-painting gun Gi) vacuum cup (1) Driling spinciie (¥) Heating torches (i) Grinders (vi) Wire brushes (vii) Arc welding tots, etc 2.2.4 Characteristics of Grippers A Gripper Must Satisfy the Following Characteristics: ‘+ Gripper must be capable of grasping, lifting and releasing the part as individual or family of parts required iby the process. ‘© Some grippers sense the presence of the part with their gripping action, ‘As far as possible weight of the tooling must be kept to a minimum value. The gripper should be simple in design, accurate in operation, economical and free of maintenance Containment of the tooling part must be assured under conditions cf maximum velocity and loss of holding power In order to accommodate overload conditions and safeguarding, gripper must be equipped with a collision sensor + The gripper for industrial robots, in a repetitive operation, ‘requires minimum gripping d There are various types of end-effectors to perform different work functions. The various types of grippers can be divided into the following major categories (Mechanical grippers (2) Vacuuen grippers or vacuum cups. (i) Magnetic grippers. (iv) Adhesive of electrostatic grippers (©) Expandable cutt 23.1 Mechanical Grippers + These are standard grippers that use mechanical fingers actuated by a mechanism to grasp an object. +The fingers called the jaws, are the appendages of the tippers that actually make contact with the object. + The fingers are either attached to the mechanism or an integral part of the machanigm, + The use of replaceable fingers allows for the wear and interchangeably, Replaceable finge's can alto be eesigned to accommodate different part models +The gripper mechanism is used to translate some form of power input to the grasping action of the fingers against the part The mechanism must be capable of opening and closing the fingers and to exert sufficient force against the part to hold it securely ‘+The input power to the mechanism is supplied from the robot and can be pneumatic, hydraulic, electric or mechanical (i. spring activated) ‘nochange Fig. 24: Interchangeable fingers in mechanical gripper + _ There are two ways of constraining the part in the gripper. @ scanned with OKEN ScannerROBOTICS (BE MECHANICAL) 1. By Physieal Constrietion of the Part within the Fingers: In this approach, the gripper fingers enclose the part 10 some extent, thus constraining the motion of the part Therefore, it is required to design contacting surtaces of the fingers to be in the approximate shape of the part geometry It is as shown in Fig. 25. ae somos Pan Gran pat Fig. 2.5: Physical constriction af the part within the finger 2. By Friction between the Fingers and the Work Part: In this approach, the fingers must apply a force that is Sufficient for friction to retain the part against gravity, acceleration and any other force that might arise during the holding portion of the work cycle. ‘©The pads attached to the fingers that make contact with the part, are generally fabricated out of a ‘material that i relative soft. This results in increase of the coefficient of friction betawen the part and the contacting finger surface It also serves to protect the Part surface from scratching of other kind of damages. = This approach results in less complicated and therefore less expensive gripper design. © Disadvantage with friction approach ~ If a force of sulfcient magnitude is applied against the part in a direction parallel to the friction surfaces of the fingers, there might be sipping of part cut of the gripper. In order to avoid this slippage. the gripper ‘must be designed to exert a force that is greater than the weight of the part: I is illustrated in Fig, 26. a Fig. 2.6 Friction approach to held a part in finger en 23.2 Closing Motions of Mechanical Gripper ROBOT GRIPPERS There are two closing motions of mechanical gripper: (@ Angularand — (i Paralel In Angular motion, the jaws move in angular sense while in paralie! motion, jaws move toward or amay from each other. It fs Mustrated in Fig. 27 and Fig, 28. be Fig. 27: Angular gripper Fig. 28 Parallel gripper 2.3.3 Vacuum Grippers Principle: Vacuum is used as the gripping force. The lifting power is a function of the degree of vacuum achieved and the size of the area on the part where the vacuum is applied Vacuum grippers work on Bernoulli's principle where suction is created by using compressed ait. The relative high vacuum i created by vacuum generator which & powered by an electric motor. Vacuum cups oF suction cups are used for lifting objects. They are made of neoprene or synthetic rubber (Characteristics of Vacuum Cups or Suction Cups: + They are simple in construction and have extreme light weight + The number, site and type of cups used will depend on the weight size, shape and type of material being handled ‘The diameter usually ranges between one to eight inches. ‘They are round oF oval in shape. +The workpiece to be handied be smooth and clean in order to form a satisfactory vacuum between the piece and the suction cup. + Flembility of the vacuum cup provides the robot with a cerain amount of compliance. Uses of Vacuum Gripper: ‘Thay can be used on curved and contoured surfaces as well 3s fiat surfaces. They are ideal for lifting fragile parts such as glass, pans of glass, lange lightweight Boxes ‘They are also used for flexible soft materials, where the ecuum cup would be made of aihard substance. @ scanned with OKEN ScannerROBOTICS (BE MECHANICAL) Working: ‘The liting power = 1 (Degree of vacuum achieved, site of area on part where the vacuum is applied Multiple-cup vacuum gripper increases the contact surface area and permit the size and weight of the workpiece to be increased. The lift capacity of the suction depends on: (i) The effective area of the cup and Gi) The negative air pressure between the cup and the object to be lifted. Fig, 2.% Venturi device fer flat surtace gripping, The relationship can be given as under Fe KP ode Ko Ae [Pa = Pass) ‘The force or lit capacity, ON) Negative pressure, (N/cm*) Total elfective area of the suction ‘Cupis) used to create the vacuum, fern’ K =A coefficient depending on “atmospheric pressure and sealing conditions, P= The atmospheric pressure Pass. = Residual pressure in vacuum cup ©The vacuum generator and venturi block (also Iinown 5 miniature vacuum pump) and twe common devices used to create suction by the use of compressed air ‘+The vacuum generator is a piston-operated or vane- driven device powered by an electric motor and is Capable of creating a relative high vacuum. +The venturi is a simple device and can be operated by ‘means of shop air pressure. ‘+ Assingle vacuum cup can produce 20 + inches of Hg vacuum from a 22 psi line. This enables the cup to Support a weight of 10 to 100 fb, depending on the sealing capacty of the parts and the desired factor of safety used. Pon ow Advantages of Vacuum Cup Grippers: They require only one surface to grasp the part They apply @ uniform pressure on the surface of the part, They need 2 relatively lightweight gripper, They are suitable to a variery of different materials ‘They have a very low cost. 23.4 Magnetic Grippers Principle: Magnetic grippers are similar in operation to vacuum grippers, however, instead of using vacuum to lift the object, they ‘employ 4 magnetic field created by an electromagnet or permanent magnet. + Magnetic grippers are employed to handle {ferromagnetic materials. + The material can be lifted in the form of a sheet or plate with an electromagnet mounted on the robot tool plate + Fig. 220 (a) shows single magnetic and Fig, 2.10 (b) dual magnetic gripper a nape rere (b) Dual magnetic gripper Fig. 220 Magnetic grippers can be categorised into ~ (a) Bectromagnets and (©) Permanent magnets (a) Electromagnetic Grippert: They are easier to control, but require a source of de power and an appropriate controller. When the part is to be released, the control unit reverses the polarity at a reduced power level before switching-off the electromagnet. This procedure acts to cancel the residual magnetism in the work piece ensuring a positive release of the part Using Maxwell's equation, the force of attraction of ‘sectromagnet can be given 36 -— ou Pe AIR Rad where, N= Number of amp-tums of coil Ac = Area of contact of an object with magnet @ scanned with OKEN Scanner