A brief overview of sensors (and why you should love them) I’m Mr. Sensitivity. . . copyright 2002: Sean Pieper, Bob Grabowski, Howie Choset Why Sensors Are Important: The world is NOT static. It can also be incredibly complex. Oh no. I thought you said the 40’th canal on sector 15. . Sometimes, the environment where a robot will be deployed is simply unknowable. . . What Is a Sensor? • Anything that detects the state of the environment. • Yep. You’ve already used sensors before in the Braitenburg lab. • Are the following sensors? – – – – positioning devices Encoders Vision Mine detectors (detector vs. sensor) Some types of Sensors: • Ladar (laser distance and ranging) – Time of flight – Phase shift • • • • • • Sonar Radar Infra-red Light sensing Heat sensing Touch sensing How to Choose a Sensor: There are four main factors to consider in choosing a sensor. 1) Cost: sensors can be expensive, especially in bulk. 2) Environment: there are many sensors that work well and predictably inside, but that choke and die outdoors. 3) Range: Most sensors work best over a certain range of distances. If something comes too close, they bottom out, and if something is too far, they cannot detect it. Choose a sensor that will detect obstacles in the range you need. 4) Field of View: depending upon what you are doing, you may want sensors that have a wider cone of detection. A wider “field of view” will cause more objects to be detected per sensor, but it also will give less information about where exactly an object is when one is detected. Creative Uses: • Sharp IR sensors are very accurate and operate well over a large range of distances proportional to the size of a Lego robot. However, they have almost no spread. This can cause a robot to miss an obstacle because of a narrow gap. One solution is to make the sensor pan. • One could also use a light sensor to detect obstacles indoors. Inside, there tend to be lights at many angles and locations. Thus, around the edges of most obstacles, a slight shadow will be cast. A light sensor could detect this shadow and thus the associated object. Warning: this could be a very fickle design. • Touch sensors can have their spread increased with large bumpers, and can be used for wall following to implement bug2. They are also dirt cheap. Example of sharp IR mounted to sweep for a wider field of view. Shadow cast indicates obstacle: one way to navigate with photo resistors. Time for some meat • Now that you understand the vital importance of sensors to constructing robust and interesting robots, you’re probably curious about what’s available. • Here’s a small sampling of commonly used sensors and their applications. . . • Hang on tight. Gas Sensor Gyro Accelerometer Pendulum Resistive Tilt Sensors Metal Detector Piezo Bend Sensor Gieger-Muller Radiation Sensor Pyroelectric Detector UV Detector Resistive Bend Sensors Digital Infrared Ranging CDS Cell Resistive Light Sensor Pressure Switch Miniature Polaroid Sensor Limit Switch Touch Switch Mechanical Tilt Sensors IR Pin Diode IR Sensor w/lens Thyristor Magnetic Sensor IR Reflection Sensor Magnetic Reed Switch IR Amplifier Sensor Hall Effect Magnetic Field Sensors Polaroid Sensor Board IRDA Transceiver Lite-On IR Remote Receiver Radio Shack Remote Receiver IR Modulator Receiver Solar Cell Compass Compass Piezo Ultrasonic Transducers Resistive Sensors Bend Sensors • Resistance = 10k to 35k • Force to produce 90deg = 5 grams • www.jameco.com = 10$ Potentiometers • Fixed Rotation Sensors • Easy to find, easy to mount Resistive Bend Sensor Potentiometer Light Sensor • Good for detecting direction/presence of light • Non-linear resistance • Slow response Cadmium Sulfide Cell Applications Sensor • Measure bend of a joint Sensors • Wall Following/Collision Detection Sensor • Weight Sensor Inputs for Resistive Sensors V1 Voltage divider: You have two resisters, one is fixed and the other varies, as well as a constant voltage R1 V Analog to Digital (pull down) R2 V2 V1 – V2 * (R2/R1+R2) = V Known unknown micro measure micro Single Pin Resistance Measurement + - Binary Threshold Comparator: if voltage at + is greater than at -, high value out Intensity Based Infrared voltage Increase in ambient light raises DC bias time voltage • Easy to implement (few components) • Works very well in controlled environments • Sensitive to ambient light time Modulated Infrared amplifier bandpass filter integrator limiter demodulator comparator Input Output 600us 600us • Insensitive to ambient light • Built in modulation decoder (typically 38-40kHz) • Used in most IR remote control units ( good for communications) • Mounted in a metal faraday cage • Cannot detect long on-pulses • Requires modulated IR signal http://www.hvwtechnologies.com http://www.digikey.com Digital Infrared Ranging Modulated IR beam Optical lenses +5v output input 1k 1k gnd position sensitive device (array of photodiodes) • Optics to covert horizontal distance to vertical distance • Insensitive to ambient light and surface type • Minimum range ~ 10cm • Beam width ~ 5deg • Designed to run on 3v -> need to protect input • Uses Shift register to exchange data (clk in = data out) • Moderately reliable for ranging Polaroid Ultrasonic Sensor • Mobile Robot Electric Measuring Tape Focus for Camera http://www.robotprojects.com/sonar/scd.htm Theory of Operation • Digital Init • Chirp – 16 high to low – -200 to 200 V • Internal Blanking • Chirp reaches object – 343.2 m/s – Temp, pressure • Echoes – Shape – Material • Returns to Xducer • Measure the time Problems • • • • • Azimuth Uncertainty Specular Reflections Multipass Highly sensitive to temperature and pressure changes Minimum Range Beam Pattern Not Gaussian!! (Naïve) Sensor Model Problem with Naïve Model Reducing Azimuth Uncertainty • • • • Pixel-Based Methods (Most Popular) Region of Constant Depth Arc Transversal Method Focusing Multiple Sensor Certainty Grid Approach Combine info with Bayes Rule (Morevac and Elfes) Arc Transversal Method • • • Uniform Distribution on Arc Consider Transversal Intersections Take the Median Mapping Example Vendors • • • • Micromint Wirz Gleason Research (Handyboard) Polaroid-oem Metal Detector Oscillator signal coupled via transformer When T2 is turned off, T3 is turned on 112kHz LC Oscillator LED will drop about 2volts Diode converts AC signal to DC ripple and applies as bias to T3 9v Signal to 5v logic +9V +5V Rpullup 9v signal + - PIC LM311 comparator A comparator can be used to convert a two-state signal to digital logic When the + voltage is above the voltage on the - pin, the output is high When the + voltage is below the - voltage, the output is low The LM311 has an open collector (you need to provide pullup resistor) This allows conversion from 9volt logic to 5volt logic Accelerometers adxl202 2-axis accelerometer • Mems technology provides precision mechanical electrical devices • ADXL202 outputs convenient PWM output whose duty cycle is proportional to acceleration • Cost about 30$ - easy to interface to PIC Accelerometer Uses • Measure tilt of arm ADXL202EB ADXL202EB • Measure Weight Analog Tilt Measurements • Pass signal through low-pass filter, then to ADC – Averages signal – Filter cutoff frequency should be < 0.1 bandwidth +5V ADXL202EB Bandwidth = 100 Hz Vi LM324 R PIC + - C 1 10 Hz 2RC 0.1mF ADC0 Vo Vi Vo t Digital Tilt Measurements • Send PWM signal directly to CCP – Set to measure pulse width – Uses a valuable microcontroller resource +5V 0.1mF ADXL202EB Bandwidth = 100 Hz Vi CCP ton accel 0.5 ton t period Pic tperiod 50% duty cycle = 0 accel Analog Velocity Measurements • Pass acceleration signal through integrator, then to ADC – Need to compensate for 2.5V offset – Need to choose RC such that Vo does not saturate – Need to periodically reset integrator to prevent overflow R R + - R ADXL202EB Bandwidth = 1000 Hz + Vi - Vo t PIC LM324 C Vi +5V ADC0 Vo Rb Threshold Measurements • Pass acceleration signal through comparator, then to input capture – Need high signal bandwidth to see pulse ADXL202EB Bandwidth = 1000 Hz R Vi LM311 comparator + - C Vo Rb PIC Rpullup +5V Vth Vo 1 100 Hz 2RC Vi Vth t