Page 133 - Toucpad robotics C11

Control Systems in Robotics: Orchestrating Accurate, Efficient, and Safe Performance

Imagine trying to drive a car with your eyes closed, or playing a sport without being able to react to the ball. It would
be impossible to achieve your goal. Similarly, a robot needs a sophisticated system to constantly monitor its actions,
compare them to what it’s supposed to do, and make corrections. This is the essence of a control system in robotics.
A control system is a set of devices and algorithms that manage, command, direct, or regulate the behaviour of a robot
or another system. Its primary objective is to ensure that the robot performs its intended task with the desired precision,
speed, and safety.

Core Components of a Robotic Control System
A typical robotic control system forms a loop, constantly working to achieve a specific outcome. It consists of three main
interconnected parts:

Sensors (The Eyes and Ears)
Role: Sensors are crucial for providing feedback about the robot’s current state and its environment. They measure
u
parameters like position, velocity, force, temperature, proximity to objects, and orientation.
Example: An encoder on a motor measures how much a robot’s joint has rotated. A vision sensor (camera) captures
u
an image of an object. A force sensor measures the pressure exerted by a gripper.
Controller (The Brain)
Role: The controller is the “decision-maker” of the control system. It receives data from the sensors, compares it to
u
the desired state (the “setpoint” or “reference”), calculates the “error” (the difference between desired and actual),
and then computes the necessary commands to reduce this error.
u Example: This could be a microcontroller, a single-board computer (like a Raspberry Pi), a dedicated industrial
robot controller, or even a powerful computer running complex Artificial Intelligence algorithms. It executes the
control algorithm.

Actuators (The Muscles)
Role: Actuators (like motors, hydraulic cylinders, or pneumatic systems) are the components that receive commands
u
from the controller and convert them into physical motion or force, thereby changing the robot’s state.
Example: If the controller determines that a robot arm needs to move to a certain position, it sends a command
u
(e.g., a voltage signal or a Pulse Width Modulation signal) to the servo motor in that joint, causing it to rotate.
Types of Control Systems
Control systems can be broadly categorised into two main types based on whether they use feedback:

Open-Loop Control Systems (Non-Feedback Systems)

Description
In an open-loop system, the control action is independent of the output. The controller sends commands to the actuators,
assuming that the robot will perform the task exactly as commanded, without checking if the desired outcome was
achieved. There is no feedback from sensors to adjust the output.
Working Principle

Command Controller Actuator Robot Action (Output).
Characteristics

u Simple Design: Easier and cheaper to implement because no sensors are needed to measure the output.
No Feedback: Does not compensate for disturbances or errors that might affect the output.
u
u Less Accurate/Reliable: If there are external disturbances (like friction, varying loads, or unexpected obstacles), the
robot will not know and cannot correct its action.

u No Automatic Correction: Cannot adapt to changing conditions.
131
Electrical and Control Systems