Page 120 - Toucpad robotics C11

Measures Linear Acceleration: Provides data on changes in velocity.
u
Compact and Robust: MEMS accelerometers are small and relatively durable.
u
Disadvantages

Confounds Gravity and Acceleration: Cannot distinguish between acceleration due to gravity and acceleration due
u
to actual motion without additional information or filtering. If a robot is falling, its accelerometer would read zero
acceleration relative to its own frame of reference.
Noise: Can be susceptible to noise from vibrations.
u
Integration Errors for Position/Velocity: Integrating acceleration twice to get position (or once for velocity) accumulates
u
errors rapidly, making it unsuitable for long-term position tracking without correction from other sensors.

Sensor Fusion: The Robot’s Holistic View
It is rare for a sophisticated robot to rely on just one type of sensor. Instead, New Age Robotics Systems employ sensor
fusion, a technique where data from multiple different sensors are combined and processed by AI algorithms to create
a more complete, accurate, and reliable understanding of the environment and the robot’s own state than any single
sensor could provide.

GPS speedometer,
RADAR
(long range) accelerometer, compass
Thermal imager

RADAR
(short range)
Sensor Fusion

LiDAR

Map V2X
Vision “seeing around corners”
Other traffic, road conditions

Example: For a mobile robot, an ultrasonic sensor might detect an obstacle’s distance, a LIDAR sensor might create
u
a detailed map of the room, and an accelerometer and gyroscope (in an Inertial Measurement Unit) might track
its own precise movements. An AI system fuses this data: the LIDAR map tells the robot where it is, the ultrasonic
sensor provides quick warnings about close objects, and the Inertial Measurement Unit refines its position tracking in
between LIDAR scans, resulting in highly robust navigation.
At last, sensors are the eyes, ears, and touch of a robot, providing the essential input for intelligent behaviour. Each type of
sensor has unique strengths and weaknesses, making their strategic combination through sensor fusion vital for building
robust, autonomous, and semi-autonomous robotic systems capable of navigating and interacting with our complex
world. Understanding these sensory inputs is as crucial as understanding the robot’s brain and muscles.
Batteries in Robotics: Powering Autonomous Operations

Just as humans need energy from food to perform tasks, robots need electrical energy to power their motors, sensors,
control units, and other electronic components. For mobile and autonomous robots that are not constantly plugged into a
power outlet, batteries serve as their portable energy reservoirs. The type of battery chosen for a robot is a critical design
decision, influencing everything from the robot’s weight and size to its operational lifespan and maintenance requirements.
Let’s explore two types of batteries commonly found in robotics applications:

118
Touchpad Robotics - XI