Robot’s Perception:
Robots are designed to imitate the human nature of sensing the surrounding and decision making, in this section we will cover some of the sensors that can be used in robotic applications to sense motion sensing, proximity, and obstacle avoidance.
- Accelerometers: our first type of motion sensors, from its name it can sense object’s acceleration in a specific direction depending on how many axes the accelerometer can sense, there are 1-D (sense acceleration in x-direction), 2-D (sense acceleration in both x- and y- directions) and 3-D (sense acceleration all three x-, y- and z- directions). The output of accelerometer is measured in terms of gravitational acceleration, an example if an accelerometer can sense maximum 1g then it can sense maximum acceleration of 9.81 m/s2. A famous example 3D accelerometers are used in is screen rotation in cellphones and tablets, they sense the direction of motion and rotate the screen.
- Gyroscopes: our second motion sensor, it can measure object’s “tilt” or rotation in a specific direction depending on how many axes the gyroscope can sense, there are 1-D, 2-D and 3-D gyroscopes (unlike the x-, y-, z- axes names used for accelerometers the three rotation motions are called “Yaw”, “Pitch” and “Roll”). Gyroscope out is measured in “degrees”, so the output voltage is equivalent to the tilt/rotation in degrees sensed.
 |
Figure 16: 3-Axes MEMS Gyroscope
(Source: Sensor Technologies for Michael J. McGarth & Cliodhna Ni Scanaill)
|
Some Accelerometers and Gyroscope chips available in the market:
- MXR9150MZ: Analog output 3-Axes accelerometer from MEMS-IC, with sensing range of +/- 5g and sensitivity of 0.15 V/g (for every 1g it senses, the output voltage will change by 0.15V).
- MXR2999EL: Analog output 2-Axes accelerometer from MEMS-IC, with sensing range of +/- 0.5g and sensitivity of 1V/g.
- ADXL335: Analog output 3-Axes accelerometer from Analog Devices, with sensing range of +/- 3g and sensitivity of 0.3V/g.
- ITG3200: Digital output 3-axes gyroscope from InvenSense, with sensing range +/- 2000 degree/second, digital output means it has integrated ADC on-chip, the output protocol is serial I2C.
- LPR5150AL: Analog output 2-Axes gyroscope from ST Microelectronics, with sensing range of +/- 1500 degrees/second.
Commercial low-cost MEMS accelerometers and gyroscopes are known for drifting over time, which mean they require calibration from time to time to fix the error due to drift, the sensor datasheet would be the best place to know this information.
- Piezoelectric shock/vibration sensors: piezoelectric materials are a type of materials that generates voltage across its surface when you exert mechanical force on it, same happens when exposed to vibrations. Though piezoelectric sensors aren’t that accurate compared to accelerometers but they’re a cheaper to measure shock/vibration.
Some of the piezoelectric sensors available in the market:
- LDT0-028K: piezoelectric vibration/shock sensor with sensitivity of 50 mV/g.
- AW2T24TEL-4A1: piezoelectric vibration/shock sensor with sensitivity of 40 mV/g.
- Micro-switches: a micro-switch isn’t a sensor, but a very soft switch that can sense slight collisions; mainly it is used as a very low-cost collision detector.
 |
Figure 17: Micro-switch
|
- Light sensors: We can sense proximity to an object by measuring reflected light from its surface. A light proximity sensor consists of a light source and a receiver that senses light intensity, Light Emitting Diodes (LED) are commonly used as the light source, while photo-resistors and photo-transistors are used for receivers, and let’s discuss the principle of operation of both types of receivers:
- Photo-resistor: also called “Light Dependent Resistor”, or LDR, is a light-controlled variable resistance, resistance changes inversely with the amount of incident light.
 |
Figure 18: Photo-resistor
|
 |
Figure 19: NPN-type photo-transistor (right), its circuit schematic (left)
|
 |
Figure 20: Simple light sensor circuit
|
- Ultrasonic sensors: our second type of proximity sensors is ultrasonic sensor; it consists of ultrasonic signal transmitter and receiver. It measures the distance to an object by measuring the time delay between transmitted and reflected signals.
 |
Figure 21: Ultrasonic sensor operation
- Infra-red sensors: IR sensors can be used for a variety of application like proximity measurement, color detection, and contactless temperature measurement. Similar to the light sensor concept of operation, infra-red, or short “IR”, sensor measures the distance/proximity to an object by measuring an emitted IR signal reflection. It consists of an IR transmitter (IR emitting diode) and IR receiver (IR phototransistor). For object detection using IR, the receiver just detects IR light reflected from the object while to detect colors the receiver has to quantify the amount of IR light reflected, dark colors reflect less IR light than bright colors.
 |
Figure 22: IR proximity sensor (left), IR color sensor (right)
|
- Laser sensors (LiDar): They measures distance by illuminating a target with a laser and analyzing the reflected light (Wikipedia). It is the most accurate and most expensive type of proximity sensors, usually used in high-end industrial and military applications like mobile robots, drones and self-driving cars.
 |
Figure 23: LIDAR-Lite 2 Laser distance measurement sensor
(source: https://www.sparkfun.com/products/13680)
|
 |
Figure 24: A LIDAR mounted on the roof-top of Google’s self-driving car.
|
We have reached the end of the second part of the series, next article will focus on the "Robot's brain" or the robot's electronic control unit, with a basic introduction to microcontrollers and some other tools that can be used for decision making and enhanced perception like machine vision, navigation, remote control, and different wireless technologies that can be used to communicate with the robot.
I hope you enjoyed the second part of the article and see in the third and last one.
--
Karim El-Rayes
January 29, 2016
Vancouver, Canada
