![[LOGO]](/workbook/skins/workbook/create_small2.png)
Sensors
Contents |
Purpose
Introduction to simple and complex sensors, how they work, the type of data they produce, and how they can be used in robotics. Refer to Chapters 7,8,9 of "The Robotics Primer".
Overview
Sensors provide a robot with the means to observe characteristics of the external environment, and internal state. Without sensors the robot would be unable to determine its current state and therefore unable to act meaningfully. The type of sensors available vary in cost, complexity of use, and precision. It is not always necessary to use to the biggest and best sensor for every task. Part of the work involved with creating a robot system is the determination of the appropriate sensors. In these exercises you will use a variety of sensors, learn how to use them and what kind of information they produce. We will start with simple range sensors such as infra-red, and progress up to very complex sensors such as lasers and camers.
Hardware
| Exercise | Command Module | Gumstix | Misc |
|---|---|---|---|
| Exercise 1 | Yes | Yes | N/A |
| Exercise 2 | Yes | Yes | N/A |
| Exercise 3 | Yes | Yes | Sonar sensor |
| Exercise 4 | Yes | Yes | IR and Sonar sensors |
| Exercise 5 | No | Yes | N/A |
| Exercise 6 | No | Yes | CMU Cam-2 |
| Exercise 7 | No | Yes | CMU Cam-2 and Sonar sensor |
Exercise1: Sensors & Levels of Processing
|
There exist a wide variety of sensors that a robot could use. Some are simple, such as bump sensors, while others are very complex, such as vision sensors. The type of sensors used on a robot depend on tasks the robot must perform, available electrical and computational power, and carrying capacity. In this exercise you will study the types of sensors equipped on the iRobot Create. |
 |
 |
Exercise2: Infra-red Sensor
|
This exercise introduces an infra-red sensor for range sensing. Infra-red sensors are very common, and can be configured in numerous ways. One common setup is a break beam sensor, where an IR emitter and sensor face each other separated by a gap. When an object comes between the sensor and emitter, a signal is produced. Another typical setup has the sensor and emitter facing the same direction. The IR produced from the emitter bouces off objects and is detected by the sensor. The time difference between emission and detection is used to calculate a distance to the object cause the reflection. This type of setup is used in this exercise. |
|
|
Exercise3: Sonar Sensor
|
Sonar is another type of sensor commonly found on robots. The same principles used in an IR range sensor apply to a sonar sensor. Sound waves are emitted from a sonar, and then detected after they bounce off objects in the environment. The time difference between emission and detection is used to determine the distance to an object. Sonar sensors typically have a longer range and wider field of view than Glossary#IR, but suffer from specular reflection. |
|
|
Exercise4: Infra-Red and Sonar wall-following
|
Wall following is a common behavior for a robot, and one that is relatively easy to implement. As the name implies, the goal is to move the robot parallel to a wall. You will use an infra-red and sonar sensor mounted on the robot to maintain a certain distance to a wall. |
|
|
Exercise5: Laser Sensor
|
While being much more expensive and bulky, laser sensors provide significantly greater precision. Futhermore, this type of laser does not suffer from specular reflection and generally has a much greater range than sonar and IR. These attributes have made lasers one of the most common types of sensors used in the research community. However, since a typical laser costs a few thousand dollars, we will use a simulated laser in this exercise. |
|
|
Exercise6: Color blob tracking
|
Camera's offer a significant source of data for a robot. Images are filled with useful information such as color, texture, movement, and 3D data when using stereo vision. However, extracting this information is computationally expensive. For this exercise you will use a CMU Cam 2, which provides hardware color blob tracking. This means the camera will process the images automatically and provide the robot with information about color blobs which appear in the image. This data will be used by the robot for tracking of colored objects. |
|
Exercise7: Person Following
|
Following a person can be a fairly complex task. A laser range finder or stereo camera could detect the legs of a person, however these two sensors are beyond the reach of this workbook. Instead we will make the job easier by marking the person with a color blob. The robot can then use a camera to follow the blob, and a sonar sensor for ranging. |
|
