Degrees of Freedom (DOF)

In mechanical systems, degrees of freedom represent the minimum number of independent coordinates required to fully specify the configuration of all bodies in the system relative to a fixed reference frame. For a single rigid body in three-dimensional space, there are six degrees of freedom: three for translational position (X, Y, Z) and three for rotational orientation (roll, pitch, yaw). For a robotic arm, the total DOF is the sum of the independent motions provided by its joints. This concept is foundational to kinematics, the study of motion without considering forces.

Each joint in a robot provides one or more degrees of freedom, constraining motion in some axes while allowing it in others.

• Revolute Joint (R): Provides 1 DOF, allowing rotational motion about a single axis (like an elbow or knee joint).
• Prismatic Joint (P): Provides 1 DOF, allowing linear sliding motion along a single axis (like a telescoping arm or piston).
• Spherical Joint (Ball Joint): Provides 3 DOF, allowing rotation about three orthogonal axes.
• Planar Joint: Provides 3 DOF (two translations and one rotation within a plane).

A robot's total DOF is calculated by summing the DOF contributed by each joint, after accounting for any redundant constraints imposed by the kinematic chain.

The arrangement of joints and links defines the robot's structure and its DOF properties.

• Serial Kinematic Chain (Open Chain): Links and joints are connected in series, like a typical industrial robot arm. The end-effector's DOF is usually equal to the total number of joint DOF. These offer a large workspace but can accumulate errors and have lower stiffness.
• Parallel Kinematic Chain (Closed Chain): The end-effector is connected to the base by multiple independent kinematic chains, like a Stewart platform (hexapod). The total DOF is often less than the sum of all joint DOF due to constraints. These provide high stiffness and precision but a smaller workspace.

The Grübler-Kutzbach criterion is a formula used to calculate the mobility (DOF) of a general mechanism, accounting for links, joints, and constraints.

DOF defines possible motions, but actuation determines which are actively controlled.

• Fully Actuated System: The number of independent actuators equals the number of DOF. Every DOF can be directly controlled.
• Underactuated System: Has fewer actuators than DOF (e.g., a quadrotor with 6 DOF but only 4 propeller actuators). Control requires dynamic coupling and is more complex.
• Kinematic Redundancy: Occurs when a robot has more DOF than required for a specific task (e.g., a 7-DOF arm performing a 6-DOF end-effector pose). Redundancy allows for optimizing secondary criteria like avoiding obstacles or minimizing joint torque.

Understanding this relationship is critical for motion planning and control system design.

In physics-based robotic simulation, accurately modeling DOF is paramount for simulation fidelity. A simulator must correctly implement:

• Joint types and limits (hard and soft stops).
• Actuator models (torque/speed curves, PID dynamics).
• Constraint solvers for closed-chain mechanisms.

Oversimplifying DOF (e.g., ignoring joint flexibility or gear backlash) is a primary contributor to the reality gap—the discrepancy between simulated and real-world robot behavior. Techniques like domain randomization vary simulated dynamics parameters, including joint friction and damping within plausible ranges, to train policies robust to these inaccuracies.

• SCARA Robot: 4 DOF (3 revolute joints in the horizontal plane, 1 prismatic joint for vertical motion).
• 6-Axis Industrial Arm: 6 DOF (typically 6 revolute joints), which is the minimum required to arbitrarily position and orient an end-effector in 3D space.
• Automobile (on a plane): 3 DOF (X, Y position, and heading orientation).
• Human Arm (from shoulder to wrist): 7 DOF (3 at shoulder, 1 at elbow, 3 at wrist), providing kinematic redundancy for dexterous manipulation.
• Quadruped Robot: A single leg may have 3 DOF (hip abduction/adduction, hip flexion/extension, knee flexion/extension). The body itself has 6 DOF in space, controlled through coordinated leg movements.

The branch of mechanics that models the motion of non-deformable objects under forces and torques. It provides the core equations of motion that govern how a system's degrees of freedom evolve over time.

• Newton-Euler Equations: Formulate the linear and angular motion of a rigid body.
• Lagrangian Dynamics: An energy-based formulation often used for complex, multi-DOF systems.
• Forward/Inverse Dynamics: Calculate motion from forces (forward) or forces from desired motion (inverse), both central to simulating robotic systems.

An algorithm within a physics engine that calculates forces to satisfy kinematic and dynamic constraints, directly limiting or defining permissible degrees of freedom.

• Joint Constraints: Enforce the motion allowed by revolute or prismatic joints, reducing the system's overall DOF.
• Contact Constraints: Prevent interpenetration and model friction between bodies, adding complex, time-varying constraints.
• Numerical Methods: Often solves a Linear Complementarity Problem (LCP) or a system of equations to resolve all constraints simultaneously within a time step.

The dual computational problems that map between a robot's joint space (defined by its degrees of freedom) and its end-effector position in Cartesian space.

• Forward Kinematics (FK): Calculates the end-effector pose from given joint angles. A deterministic function for serial chains.
• Inverse Kinematics (IK): Calculates the joint angles required to achieve a desired end-effector pose. Often underdetermined (infinite solutions) for high-DOF arms, requiring optimization.
• Jacobian Matrix: A linear mapping between joint velocities (DOF velocities) and end-effector velocity, critical for IK and force control.

File formats that formally describe a robot's kinematic tree, thereby defining its degrees of freedom for simulation and control.

• URDF (Unified Robot Description Format): An XML format defining robot links (rigid bodies) and joints (revolute, prismatic, fixed). Each non-fixed joint adds one DOF.
• SDF (Simulation Description Format): A more comprehensive format than URDF, capable of describing entire simulated worlds with nested models, lights, and sensors, while still specifying joint-based DOF.
• Usage: These files are parsed by simulators like Gazebo, PyBullet, and MuJoCo to instantiate a simulated robot with the correct dynamic properties.

The simulation of a motor or actuator that drives a degree of freedom. It defines the limits and dynamics between a control command and the resulting joint motion or force.

• Ideal vs. Realistic Models: An ideal model instantly achieves commanded position/torque. A realistic model includes saturation limits (max torque/velocity), response latency, and dynamics (e.g., modeled as a PID controller).
• Control Modes: Actuators can be simulated in position-control, velocity-control, or torque-control modes, each affecting how the DOF is commanded.
• Importance for Sim2Real: High-fidelity actuator models are critical for bridging the reality gap, as physical motor dynamics are a primary source of discrepancy.

Also known as the Articulated Body Algorithm, this is an efficient O(n) computational method for performing forward and inverse dynamics calculations on serial-chain robotic manipulators, where n is the number of degrees of freedom.

• Computational Efficiency: Dramatically faster than O(n³) naive methods for robots with many DOF (e.g., humanoid robots).
• Recursive Formulation: Propagates calculations through the kinematic tree from base to tip (forward pass) and back (backward pass).
• Industry Standard: The algorithm of choice in high-performance physics engines like MuJoCo for computing the motion of complex articulated systems.