Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Lecture FIVE Sensors 5 Sensors Sensors are the robot’s contact with the outside world and used to sense or measure the robot’s environment or its own internal parameters such as temperature, force, luminance, resistance to touch, weight, size, etc. These might include active and passive IR (infra-red) sensors; sound and voice sensors; ultrasonic range sensors, positional encoders on arm joints, head and wheels; compasses, navigational and GPS sensors; active and passive light and laser sensors; a number of bumper switches; and sensors to detect acceleration, turning, tilt, odour detection, magnetic fields, ionizing radiation, temperature, tactile, force, torque, video, and numerous other types. We will discuss all these sensors in four categories. These are Range, Proximity, Touch, and Force-torque sensing We classify sensors using two important functional axes: Proprioceptive / exteroceptive and Passive/active. Proprioceptive sensors measure values internal to the robot; for example, motor speed, wheel load, robot arm joint angles, battery voltage. Exteroceptive sensors acquire information from the robot’s environment; for example, distance measurements, light intensity, sound amplitude. Hence exteroceptive sensor measurements are interpreted by the robot in order to extract meaningful environmental features. Passive sensors measure ambient environmental energy entering the sensor. Examples of passive sensors include temperature probes, microphones, and CCD or CMOS cameras. Active sensors emit energy into the environment, then measure the environmental reaction. Because active sensors can manage more controlled interactions with the environment, they often achieve superior performance. However, active sensing introduces several risks: the outbound energy may affect the very characteristics that the sensor is attempting to measure. 1 5.1 Range sensors A range sensor measures the distance from a reference point to an object in the field of operation. In such sensors time-of-flight concept is used in which distance is estimated based on the time elapsed between the transmission of signal and return of reflection. A sensor consists of two parts: a transducer to produce wave energy, and an aperture or antenna to radiate or receive such energy. However these may be integrated into a single component. • The energy is launched as a wave across an aperture or antenna • It propagates through the atmosphere until it meets a reflector • A proportion is backscattered to the receiver and gathered by an antenna • The energy is converted back into an electrical signal and stored in a computer Among the most common range sensors are: • Infrared (IR), • Sonar, and • Laser sensors Infrared (IR) sensors IR sensors are non-contact sensors used to detect obstacles. They operate by emitting an infrared light and detecting reflection from objects in front of the robot. IR sensor measurements mainly depend on the surface and colour of the object. For example, black objects are invisible to IR sensors. Since the IR signal is inversely proportional to distance, IR sensors are inherently short range sensors. IR sensors are usually divided into two basic types: the passive IR sensors that emit no IR radiation and the active types that emit an IR beam that is again detected by reflection. The active IR sensors generally use an IR LED emitting an invisible beam that is, in turn, picked up as a reflected spot on a wall or object by a photo transistor. Some IR sensor products are shown in Figure 5.1. Fig. 5.1: Infra-red light sensors 2 Sonar sensors Sonar sensors emit a short powerful signal and receive the reflection off objects ahead of the sensor. The distance of the object is calculated from the travel time of the signal and the speed of sound. The general principle of sonar sensor is shown in Figure 5.2. Fig. 5.2: The principles of first return sonar using a threshold detector. The counter stops when the signal received exceeds a pre-set threshold. The time of flight can be read directly from the counter and converted to distance. The system could be modified to use a threshold which decreases with time over the receiving period to account for the attenuation of the sonar signal with range. An example of different noises of sonar sensors is illustrated in Figure 5.3. It can be seen that the basic shape for the rough surface varies significantly with bearing angle. The basic shape of a smooth surface image does not vary significantly, but there is significant change in the amplitude of the peak with angle of incidence (arising from the angular variation of the radiation pattern). Uncertainties in sonar sensors Various uncertainties are associated with the readings of sonar sensors. The sensitivity of a sonar sensor is cone-shaped. As a result the object distance can be anywhere in the sonar cone. Also the accuracy of the object distance is a function of the width of the sonar beam pattern. Sonar sensor gives erroneous readings due to specular reflection. Specular reflection occurs when sonar beam hits a smooth surface at a shallow angle and therefore does not reflect back to sensor but outward. 3 Fig. 5.3 Sonar images of surfaces: (a) image of a smooth surface at zero bearing; (b) image of a smooth surface obtained at the same distance but a bearing angle of about 28° (the second small peak in plot (b) is due to a further reflector in the line of sight); (c) image of a rough surface at zero bearing angle; and (d) image of a rough surface obtained at the same position but bearing angle of about 28°. Laser range finders Laser range finders are particularly very common in mobile robots to measure the distance, velocity, and acceleration of objects. A short light signal is sent out and the reflection off object is detected to measure the elapsed time. Shorter wavelength reduces the specular reflection. The very inexpensive diode lasers available as pointers and power tool line generators make great robot add-ons. Fig. 5.4: High-end-lesser and obstacle detectors 4 Triangulation-based active ranging Triangulation-based ranging sensors use geometric properties manifest in their measuring strategy to establish distance readings to objects. The simplest class of triangulation-based rangers are active because they project a known light pattern (e.g., a point, a line, or a texture) onto the environment. The reflection of the known pattern is captured by a receiver and, together with known geometric values, the system can use simple triangulation to establish range measurements. If the receiver measures the position of the reflection along a single axis, we call the sensor an optical triangulation sensor in 1D. If the receiver measures the position of the reflection along two orthogonal axes, we call the sensor a structured light sensor. These two sensor types are described in the two sections below. Optical triangulation (1D sensor) The principle of optical triangulation in 1D is straightforward, as depicted in figure 5.5. A collimated beam (e.g., focused infrared LED, laser beam) is transmitted toward the target. The reflected light is collected by a lens and projected onto a position-sensitive device (PSD) or linear camera. Given the geometry of Figure 5.5, the distance is given by equation (5.1): D f L x (5.1) The distance is proportional to 1 / x ; therefore the sensor resolution is best for close objects and becomes poor at a distance. Sensors based on this principle are used in range sensing up to 1 or 2 m, but also in high-precision industrial measurements with resolutions far below 1 µm. Optical triangulation devices can provide relatively high accuracy with very good resolution (for close objects). However, the operating range of such a device is normally fairly limited by geometry. Fig 5.5: Principle of 1D laser triangulation. 5 5.2 Proximity sensors Proximity sensors generally have a binary output which indicates the presence of an object within a specified distance interval. They are used in robotics for grasping or avoiding obstacles. Among the most widely used proximity sensors are: • Inductive sensors, • Hall-effect sensors, • Capacitive sensors, • Ultrasonic sensors, and • Optical proximity sensors. Inductive sensors are based on a charge of inductance due to presence of a ferromagnetic metallic object. The voltage waveform observed at the output of the coil provides an effective means for proximity sensing. Hall-effect sensors are based on Lorentz force which acts on a charged particle travelling through a magnetic field. Bringing a ferromagnetic material close to the semiconductormagnetic device would decrease the strength of the magnetic field, thus reducing the Lorentz force and the voltage across the semiconductor. This drop in voltage is the key to sensing proximity with Hall-effect sensors. Capacitive sensors are potentially capable of detecting all solid and liquid materials. Capacitive sensors are based on detecting a charge in capacitance induced by a surface that is brought near the sensing element. The sensing element is a capacitor composed of a sensitive electrode and a reference electrode. Typically, these sensors are operated in a binary mode so that a change in the capacitance greater than a threshold indicates the presence of an object. Ultrasonic sensors reduce the dependence of material being sensed. The basic element is an electro-acoustic transducer of piezoelectric ceramic type. The same transducer is used for both transmitting and receiving. The housing is designed so that it produces a narrow acoustic beam for efficient energy transfer and signal direction. Proximity of an object is detected by analysing the waveforms of the both transmission and detection of acoustic energy signals. Fig. 5.6: Ultra-sonic rage finders 6 Optical proximity sensors detect the proximity of an object by its influence on a propagating wave as it travels from a transmitter to a receiver. This sensor consists of a solid-state LED, which acts as a transmitter of an infrared light, and solid-state photodiode which acts as the receiver. The cones of light formed by focusing the source and the detector on the same plane intersect in a long, pencil-like volume. This volume defines the field of operation of the sensor. A reflective surface that intersects the volume is illuminated by the source and seen by the receiver. 5.3 Force and Torque sensors Force and torque sensors are used for measuring the reaction forces developed at the joints. A joint sensor measures the Cartesian components of force and torque acting on a robot joint. Most wrist sensors function as transducers for transforming forces and moments exerted at the hand into measurable deflections or displacements at the wrist. They consist of strain gauges that measure the deflection of the mechanical structure due to external forces. Shaft encoders are used to measure the movement of robot’s motors for both translational and rotational movement. Positional encoders are probably the second most popular sensor on a robot. Most experimental robots do not have arms and do not use positional encoders to determine the positions of an arm’s different joints. They do use shaft encoders on the wheels or motor shafts to determine the number of revolutions of the wheels and thus, the distance travelled by the wheels. These encoders can use electrical contacts, magnetic Hall-effect detectors, or the more popular optical path broken by rotating teeth or opaque and clear graphics etched on a wheel. Absolute encoders output a binary word for each incremental position and are complex and expensive. Incremental encoders provide a pulse for each increment of shaft movement. The use of two optical channels enables the determination of the direction of rotation. 5.4 Touch sensors Touch sensors are used in robots to obtain information associated with the contact between a manipulator hand objects in the workspace. Touch sensors can be subdivided into two groups: binary and analogue. Binary sensors are basically contact devices such as micro-switches to detect presence of an object in between end-effectors. On the other hand, analogue sensors are compliant devices that output a signal proportional to force. During the past few years, considerable effort has been made to the development of tactile sensing arrays capable of producing touch information over a wider area of robot finger or hand. Using such several approaches significant progress has been made in the construction of artificial skin. 5.5 Vision sensors Vision is our most powerful sense. It provides us with an enormous amount of information about the environment and enables rich, intelligent interaction in dynamic environments. The first step in this process is the creation of sensing devices that capture the same raw information light that the human vision system uses. Some examples are shown in Figure 5.7. These sensors have specific limitations in performance when compared to the human eye. 7 Fig. 5.7: Camera and vision sensors Two current technologies for vision sensors will be discussed: CCD and CMOS. CCD technology The charged coupled device is the most popular basic ingredient of robotic vision systems today. The CCD chip, shown in Figure 5.8, is an array of light-sensitive picture elements, or pixels, usually with between 20,000 and several million pixels total. Each pixel can be thought of as a light-sensitive, discharging capacitor that is 5 to 25 µm in size. First, the capacitors of all pixels are charged fully, and then the integration period begins. As photons of light strike each pixel, they liberate electrons, which are captured by electric fields and retained at the pixel. Over time, each pixel accumulates a varying level of charge based on the total number of photons that have struck it. After the integration period is complete, the relative charges of all pixels need to be frozen and read. In a CCD, the reading process is performed at one corner of the CCD chip. The bottom row of pixel charges is transported to this corner and read, then the rows above shift down and the process is repeated. This means that each charge must be transported across the chip, and it is critical that the value be preserved. This requires specialized control circuitry and custom fabrication techniques to ensure the stability of transported charges. The photodiodes used in CCD chips (and CMOS chips as well) are not equally sensitive to all frequencies of light. They are sensitive to light between 400 and 1000 nm wavelength. It is important to remember that photodiodes are less sensitive to the ultraviolet end of the spectrum (e.g., blue) and are overly sensitive to the infrared portion (e.g., heat). 8 Fig 5.8: 2048x2048 CCD array There are two common approaches for creating colour images. If the pixels on the CCD chip are grouped into 2 x 2 sets of four, then red, green, and blue dyes can be applied to a colour filter so that each individual pixel receives only light of one colour. Normally, two pixels measure green while one pixel each measures red and blue light intensity. CMOS Technology The complementary metal oxide semiconductor chip is a significant departure from the CCD. It too has an array of pixels, but located alongside each pixel are several transistors specific to that pixel. Just as in CCD chips, all of the pixels accumulate charge during the integration period. During the data collection step, the CMOS takes a new approach: the pixel-specific circuitry next to every pixel measures and amplifies the pixel’s signal, all in parallel for every pixel in the array. Using more traditional traces from general semiconductor chips, the resulting pixel values are all carried to their destinations. The CMOS chip is so much simpler that it consumes significantly less power; incredibly, it operates with a power consumption that is one-hundredth the power consumption of a CCD chip. In a mobile robot, power is a scarce resource and therefore this is an important advantage. A commonly available low-cost CMOS camera is shown in Figure 5.8. Fig. 5.8: A commercially available, low-cost CMOS camera with lens attached. 9 Visual ranging sensors As we have seen earlier, a number of sensors are popular in robotics explicitly for their ability to recover depth estimates. A fundamental problem with visual images makes range finding relatively difficult. Any vision chip collapses the 3D world into a 2D image plane, thereby losing depth information. If one can make strong assumptions regarding the size of objects in the world, or their particular colour and reflectance, then one can directly interpret the appearance of the 2D image to recover depth. The general solution is to recover depth by looking at several images of the scene to gain more information. An alternative is to create different images, not by changing the viewpoint, but by changing the camera geometry, such as the focus position or lens iris. This is the fundamental idea behind depth from focus and depth from defocus techniques. The basic formula governing image formation relates the distance of the object from the Lens d to the distance e from the lens to the focal point, based on the focal length f of the lens: 1 1 1 f d e (5.2) Example The formation of an image using a camera with a thin convex lens is shown in Figure 5.9. An object with a height of 0.5m is placed at a distance of 1m from the lens with a focal length of 35mm. What will be the height of the image? Lens Objec t ho O Image f hi do di Figure 5.9: Formation of an image with a convex thin lens The distance of the image d i can be calculated using the well-known formula di 1 1 1 do di f f *do 0.035 = = 0.03627 m 0.965 do f Once the distance of the image d i is known, the height of the image can be calculated using 10 the relationship as follows d o ho di hi hi ho di 0.03627 0.5 0.018135 m do 1.0 5.6 Machine olfaction (electronic nose) Machine olfaction is the automated simulation of the sense of smell. It is an emerging requirement of modern robotics where robots or other automated systems are needed to measure the existence of a particular chemical concentration in air. This technology is still in the early stages of development, but it promises many applications, such as: quality control in food processing (e.g. taints, bacterial spoilage) detection and diagnosis in medicine detection of drugs, explosives and dangerous or illegal substances military and law enforcement (e.g. chemical warfare agents) disaster response (e.g. toxic industrial chemicals) environmental monitoring (e.g. pollutants) Pattern analysis constitutes a critical building block in the development of gas sensor array instruments capable of detecting, identifying, and measuring volatile compounds, a technology that has been proposed as an artificial substitute for the human olfactory system. Some pattern recognition problems in machine olfaction such as odor classification and odor localization can be solved by using time series kernel methods. There are three basic detection techniques using: Conductive-polymer odour sensors (poly-pyrrole) Tin-oxide gas sensors Quartz-crystal micro-balance sensor They generally comprise an array of sensors of some type, the electronics to interrogate those sensors and produce the digital signals, and finally the data processing and user interface software. Conventional electronic noses are not analytical instruments in the classical sense and very few claim to be able to quantify an odour. These instruments are first ‘trained’ with the target odour and then used to ‘recognise’ smells so that future samples can be identified as ‘good’ or ‘bad’ smells. Electronic noses have been demonstrated to discriminate between odours and volatiles from a wide range of sources. The list below shows just some of the typical applications for electronic nose technology – many are backed by research studies and published technical papers. 11 References 1. Bekey, G.A. (2005). Autonomous Robots: From Biological Inspiration to Implementation and Control, The MIT Press. 2. Dowling, K. (1997). Limbless Locomotion: Learning to Crawl with a Snake Robot, PhD Thesis, Robotic Institute, Carnegie Mellon University, USA. 3. Hirose, S. (1993). Biologically Inspired Robots (Snake-like Locomotor and Manipulator, Oxford University Press. 4. Miller, G.S.P (2002). Snake Robot for Search and Rescue, in Neurotechnology for Biomimetic Robots, Chapter 13, eds. J. Ayers, J.L. Davis, A. Rudolph, A Bradford Book, The MIT Press. 5. Schilling, R.J. (1990). Fundamentals of Robotics: Analysis and Control, Prentice Hall, Englewood Cliffs, New Jersey. 6. Siegwart, R. and Nourbakhsh, I. R. (2004). Introduction to Autonomous Mobile Robot, A Bradford Book, The MIT Press. 7. Shuzhi, S.G. and Lewis, F.L. (2006). Autonomous Mobile Robots: Sensing, Control, Decision Making and Applications, Taylor and Francis, Boca Raton, London, NY. 12