Skip to main content

Technical Glossary

A comprehensive glossary of robotics and AI terms used throughout the textbook.


A

Action Space [SOURCE NEEDED: Definition and examples]

Actuator A component that converts electrical, hydraulic, or pneumatic energy into mechanical motion. Examples: DC motors, servo motors, linear actuators.

ADAS (Advanced Driver-Assistance Systems) [SOURCE NEEDED: Definition]


B

Behavior Cloning [SOURCE NEEDED: Imitation learning technique]

Bipedal Locomotion [SOURCE NEEDED: Two-legged walking]


C

CLIP (Contrastive Language-Image Pre-training) [SOURCE NEEDED: Vision-language model explanation]

CUDA (Compute Unified Device Architecture) [SOURCE NEEDED: NVIDIA GPU programming platform]


D

DDS (Data Distribution Service) The middleware protocol used by ROS 2 for real-time, scalable, and reliable communication between nodes.

DOF (Degrees of Freedom) [SOURCE NEEDED: Definition with robotics examples]

Domain Randomization [SOURCE NEEDED: Sim-to-real transfer technique]


E

Embodied AI Artificial intelligence systems that interact with the physical world through sensors and actuators, as opposed to disembodied AI that operates purely in digital spaces.

End-Effector [SOURCE NEEDED: Robot arm tool/gripper definition]


F

Forward Kinematics [SOURCE NEEDED: Computing end-effector pose from joint angles]


G

Gazebo An open-source 3D robotics simulator with physics engines, sensor simulation, and ROS integration.

GPU (Graphics Processing Unit) [SOURCE NEEDED: Role in robotics and AI]


H

Humanoid Robot [SOURCE NEEDED: Anthropomorphic robot definition]


I

IMU (Inertial Measurement Unit) A sensor that measures acceleration, angular velocity, and sometimes magnetic field, used for estimating orientation and motion.

Inverse Kinematics [SOURCE NEEDED: Computing joint angles from desired end-effector pose]

Isaac Gym [SOURCE NEEDED: NVIDIA's massively parallel RL training platform]

Isaac Sim [SOURCE NEEDED: NVIDIA's photorealistic robot simulator]


J

Jacobian [SOURCE NEEDED: Matrix relating joint velocities to end-effector velocities]

Jetson [SOURCE NEEDED: NVIDIA's edge AI computing platform]


K

KaTeX A fast math typesetting library for the web, used to render LaTeX equations in this textbook.

Kinematics [SOURCE NEEDED: Study of motion without considering forces]


L

LiDAR (Light Detection and Ranging) A sensor that uses laser pulses to measure distances, creating 3D point clouds of the environment.

LLM (Large Language Model) [SOURCE NEEDED: Definition with robotics applications]


M

Manipulation [SOURCE NEEDED: Robot grasping and object handling]

Motion Planning [SOURCE NEEDED: Computing collision-free paths]


N

Navigation Stack [SOURCE NEEDED: ROS 2 Nav2 components]


O

Observation Space [SOURCE NEEDED: RL state representation]

Odometry [SOURCE NEEDED: Estimating position from motion sensors]


P

Perception-Action Loop The fundamental cycle in robotics: sense environment → process information → decide action → execute → sense again.

PID Controller [SOURCE NEEDED: Proportional-Integral-Derivative control]

Point Cloud [SOURCE NEEDED: 3D representation from depth sensors]

PPO (Proximal Policy Optimization) [SOURCE NEEDED: RL algorithm]


Q

QoS (Quality of Service) [SOURCE NEEDED: ROS 2 communication reliability policies]


R

Reinforcement Learning (RL) [SOURCE NEEDED: Learning by trial and error]

RGB-D Camera [SOURCE NEEDED: Camera with color and depth]

ROS (Robot Operating System) An open-source middleware framework for robot software development.

ROS 2 The second generation of ROS, with improved real-time capabilities, security, and DDS-based communication.

RViz [SOURCE NEEDED: ROS visualization tool]


S

SDF (Simulation Description Format) [SOURCE NEEDED: Gazebo's robot/world description format]

Sensor Fusion [SOURCE NEEDED: Combining multiple sensor inputs]

Servo Motor [SOURCE NEEDED: Motor with position feedback]

SLAM (Simultaneous Localization and Mapping) [SOURCE NEEDED: Building maps while localizing]


T

TensorRT [SOURCE NEEDED: NVIDIA's inference optimization engine]

Trajectory [SOURCE NEEDED: Time-parameterized path]


U

URDF (Unified Robot Description Format) An XML-based format used in ROS to describe robot kinematics, dynamics, and visual properties.


V

VLA (Vision-Language-Action) Models that integrate computer vision, natural language understanding, and robotic control to enable robots to perform tasks specified in natural language.

ViT (Vision Transformer) [SOURCE NEEDED: Transformer-based vision model]


W

Whole-Body Control [SOURCE NEEDED: Coordinating all DOFs of humanoid]


X

Xacro (XML Macros) [SOURCE NEEDED: Macros for URDF files]


Y

YOLO (You Only Look Once) [SOURCE NEEDED: Real-time object detection]


Z

Zero-Shot Learning [SOURCE NEEDED: Generalizing to unseen classes]


Return to Course Home