liveBook · Manning

Sensor

As explained in Robotics for Programmers

A device that measures aspects of a robot or its environment and provides data representing those measurements.

A sensor is a robot’s input device: it measures a physical quantity and exposes data representing that measurement to robot software. Sensors connect the physical environment to computation, while sensing software gives their raw output meaning. section 1.2.2 section 1.2.3

Common sensor input-output pairs include:

  • A temperature sensor converts heat into a temperature expressed in kelvin, Celsius, or Fahrenheit. section 1.2.2
  • A camera sensor converts light into a two-dimensional array of spatial light intensities. section 1.2.2
  • A joint position sensor converts relative body rotation into angular measurements such as increments, degrees, or radians. section 1.2.2
  • Sensor hardware can contain firmware that processes measurements; users generally interact with it as a black box through a known API. section 1.2.2

Measurements are imperfect: sensor noise makes values fluctuate around the real value, and every sensor has limited accuracy. Accuracy, cost, and size therefore create practical tradeoffs. Hardware physics is usually hidden behind the sensor interface; processing belongs in sensing when it extracts information without considering the robot’s objectives. section 1.2.2 section 1.2.3

Figure 1.1 The conceptual structure of a robot system forms a continuous loop across six essential parts. Sensors acquire raw data from the environment, which sensing processes into meaningful information. Planning uses this information to calculate the next action, which acting translates into physical commands for actuators that ultimately alter the environment.
CH01 F01 bihlmaier mm abstract

Trace the sensor’s role in the continuous robot loop: the environment produces raw data, sensing interprets it, planning selects an action, and acting drives actuators that change the environment. section 1.2

ReBo(o)t

A minimal robot for automatically rebooting a crashed computer uses one light sensor and one actuator. Its simple planning algorithm responds to the sensor input, illustrating how a physical measurement can participate in a complete robot system. section 3.1

How Robotics for Programmers covers it

How deep it goes
Defines sensor interfaces, examples, noise, and accuracy tradeoffs, but defers physical conversion principles and detailed sensor types.
Examples
Uses temperature sensors, camera sensors, joint position sensors, and ReBo(o)t’s single light sensor.
What it assumes
Assumes basic programming familiarity while building robotics understanding from sensor inputs, outputs, and software interfaces.
What you can do after
Evaluate sensor outputs as imperfect data and place their processing correctly within a robot control loop.
How it connects
Connects sensors to environment, sensing, planning, actuators, robot models, and forward kinematics.

Also explained in

Compare how 3 books explain sensor →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Robotics for Programmers. The section links point to the parts of the book it draws on.