Glossary
Key Concepts in Physical AI and Humanoid Robotics
A
Action (ROS 2)
A communication pattern in ROS 2 for long-running, goal-oriented tasks that provide continuous feedback and can be canceled. Used for complex behaviors like navigation or manipulation.
Actuator
A mechanical device that converts energy (electrical, hydraulic, pneumatic) into motion. In robotics, actuators control joint movements and end-effector actions.
B
Bipedal Locomotion
Walking or running on two legs. Requires sophisticated control to maintain balance and stability, making it one of the most challenging aspects of humanoid robotics.
C
Center of Mass (CoM)
The average position of all mass in a system. For humanoid robots, controlling the CoM is critical for maintaining balance during locomotion and manipulation.
Cognitive Robotics
The field of robotics focused on developing robots with cognitive capabilities such as reasoning, learning, planning, and natural interaction with humans.
Costmap
A grid-based representation of the environment used in navigation, where each cell contains information about obstacles and traversability.
D
Digital Twin
A virtual replica of a physical system that mirrors its state, behavior, and environment. In robotics, digital twins enable safe testing and rapid iteration in simulation.
Domain Randomization
A technique for improving sim-to-real transfer by training AI models on simulated environments with randomized parameters (lighting, textures, physics).
Degrees of Freedom (DOF)
The number of independent parameters that define the configuration of a mechanical system. A humanoid robot typically has 20-30 DOF.
E
Embodied Intelligence
The concept that an intelligent agent's cognitive capabilities are deeply intertwined with its physical body and interactions with the environment.
End-Effector
The device at the end of a robotic arm or manipulator that interacts with the environment (e.g., gripper, hand, tool).
F
Forward Kinematics
The process of calculating the position and orientation of a robot's end-effector given its joint angles.
G
Gait
A rhythmic pattern of limb movements used for locomotion. Common gaits include walking, running, and hopping.
Gazebo
An open-source 3D robotics simulator with accurate physics simulation and extensive ROS integration.
GPT (Generative Pre-trained Transformer)
A large language model architecture used for natural language understanding, generation, and task planning in conversational robotics.
H
Humanoid Dynamics
The study of forces and motions governing humanoid robots, including balance, stability, and whole-body coordination.
I
IMU (Inertial Measurement Unit)
A sensor that measures orientation, angular velocity, and linear acceleration using accelerometers and gyroscopes.
Inverse Kinematics
The process of calculating the joint angles required to place a robot's end-effector at a desired position and orientation.
Isaac Sim
NVIDIA's photorealistic robotics simulation platform built on Omniverse, designed for AI training and synthetic data generation.
J
Joint
A connection between two links in a robot that allows relative motion. Common types include revolute (rotational) and prismatic (linear).
K
Kinematics
The study of motion without considering forces. In robotics, kinematics describes the relationship between joint angles and end-effector positions.
L
LiDAR (Light Detection and Ranging)
A sensor that measures distances by illuminating targets with laser light and measuring the reflection with a sensor.
Localization
The process of determining a robot's position and orientation within an environment.
M
Manipulation
The act of physically interacting with objects in the environment, typically using robotic arms or grippers.
Mapping
The process of building a representation of an environment, often as a 2D occupancy grid or 3D point cloud.
Model Predictive Control (MPC)
A control strategy that optimizes a sequence of future control actions by predicting system behavior over a finite time horizon.
Multimodal Fusion
Combining information from multiple sensor modalities (vision, audio, touch) to improve perception and decision-making.
N
Nav2 (Navigation 2)
The second generation of ROS navigation software, providing flexible path planning and control for mobile robots.
Node (ROS 2)
An executable process in ROS 2 that performs a specific task and communicates with other nodes via topics, services, or actions.
O
Odometry
The estimation of a robot's position and orientation over time based on motion sensor data (wheel encoders, IMU, visual features).
P
Perception
The process by which robots acquire, interpret, and understand information from their environment using sensors.
Physical AI
The study and development of intelligent systems that perceive, reason, and act within physical environments, as opposed to purely digital AI.
PhysX
NVIDIA's physics simulation engine used in Isaac Sim and other robotics platforms for accurate rigid body dynamics.
R
RGB-D Camera
A camera that captures both color (RGB) and depth (D) information, enabling 3D perception.
ROS 2 (Robot Operating System 2)
A flexible framework for writing robot software, providing tools and libraries for building distributed robotic systems.
Reinforcement Learning (RL)
A machine learning paradigm where an agent learns to make decisions by interacting with an environment and receiving rewards.
S
Service (ROS 2)
A synchronous request-response communication pattern in ROS 2, used for one-time operations like configuration or queries.
SLAM (Simultaneous Localization and Mapping)
A computational problem where a robot simultaneously constructs a map of an unknown environment while localizing itself within that map.
Synthetic Data
Artificially generated data created in simulation, used to train AI models without requiring real-world data collection.
T
Topic (ROS 2)
An asynchronous publish-subscribe communication channel in ROS 2, used for continuous data streams like sensor readings.
U
URDF (Unified Robot Description Format)
An XML format for describing robot kinematics, dynamics, and visual properties in ROS.
Unity
A real-time 3D development platform used for high-fidelity robot visualization and human-robot interaction studies.
V
VLA (Vision-Language-Action)
A system architecture that integrates visual perception, natural language understanding, and physical action generation for cognitive robotics.
VSLAM (Visual SLAM)
SLAM using camera images as the primary sensor input, enabling localization and mapping through visual features.
W
Whisper
OpenAI's automatic speech recognition (ASR) system, used for converting human speech to text in voice-controlled robotics.
Whole-Body Control
A control strategy that coordinates all joints of a humanoid robot simultaneously to achieve desired movements while maintaining balance.
X
XACRO (XML Macros)
An extension of URDF that adds programmable features like macros, properties, and mathematical expressions for more maintainable robot descriptions.
Z
Zero Moment Point (ZMP)
A point on the ground where the total moment from gravity and inertia forces equals zero. Critical for bipedal balance control.
Navigation:
← Hardware Architecture | Bibliography →