Denavit–Hartenberg parameters

Denavit–Hartenberg parameters
Ethan Tira-Thompson · CC BY-SA 3.0 · source
Name	Denavit–Hartenberg parameters
Related	Jacobian (robotics), Forward kinematics, Inverse kinematics

Denavit–Hartenberg parameters are a standardized convention for describing the relative geometry of adjacent links in articulated manipulators, introduced to simplify the derivation of kinematic equations for serial robots. The convention reduces the number of parameters needed to specify the relative position and orientation of link coordinate frames and is widely used in robotics, controls, and automation. It underpins many formulations of forward kinematics, inverse kinematics, and the computation of the Jacobian (robotics) for manipulators designed by organizations such as Unimation, KUKA, ABB Robotics and research groups at Massachusetts Institute of Technology, Stanford University, and Carnegie Mellon University.

Introduction

The Denavit–Hartenberg convention originated from work by Jacques Denavit and Richard Hartenberg and was popularized through textbooks and courses at institutions like Massachusetts Institute of Technology and Stanford University. It provides a systematic method to assign coordinate frames to links of serial chains such as those in industrial robots from Unimation or service robots tested at Carnegie Mellon University. The approach interfaces with algorithms developed at NASA and research done at ETH Zurich and Tokyo Institute of Technology for analyzing manipulators used in projects at Jet Propulsion Laboratory and European Space Agency.

Definition and Convention

The Denavit–Hartenberg convention defines four parameters per link: two distances and two angles that relate successive coordinate frames. The parameters are typically denoted a_i, alpha_i, d_i, and theta_i in robotics textbooks authored by scholars at Massachusetts Institute of Technology, University of California, Berkeley, and Georgia Institute of Technology. The assignment rules constrain how axes are chosen, which simplifies kinematic chain description used in analyses at MIT Media Lab, Carnegie Mellon University robotics institute, and industrial labs at Siemens and General Motors.

Mathematical Formulation

Each link transformation is expressed by a homogeneous transformation matrix built from the four Denavit–Hartenberg parameters; this matrix composes rotations and translations to move between link frames. The formulation appears in canonical texts from Springer Science+Business Media and Elsevier and is used in computational tools developed at MathWorks (for MATLAB) and open-source projects hosted by developers at GitHub and research groups at ETH Zurich. The product of successive link matrices yields the manipulator's forward kinematics, a technique employed in control systems designed at NASA Jet Propulsion Laboratory and DARPA-funded robotics programs.

Applications in Robot Kinematics

Denavit–Hartenberg parameters are applied to derive forward and inverse kinematics, compute Jacobians, and perform dynamics modeling for manipulators used by companies like ABB Robotics, KUKA, and laboratories at Stanford University and Carnegie Mellon University. They are central to motion planning algorithms developed in projects at Google DeepMind-adjacent research groups, manipulation research at OpenAI collaborations, and industrial automation standards referenced by International Organization for Standardization. Educational courses at institutions such as Massachusetts Institute of Technology, Imperial College London, and University of Tokyo teach DH-based methods before introducing alternatives.

Limitations and Alternatives

Despite its wide use, the Denavit–Hartenberg convention has limitations: frame assignment can be non-intuitive for complex topologies, singular configurations, and parallel joint axes common in mechanisms studied at Pratt Institute and University of Pennsylvania. Alternatives and extensions include modified Denavit–Hartenberg conventions, product of exponentials (PoE) formulation popularized by researchers at University of California, Berkeley and Cornell University, and screw theory approaches taught at Imperial College London and used in work at ETH Zurich. These alternatives are often presented in literature from Springer and discussed in symposia such as those organized by IEEE and IFAC.

Example Calculations and Worked Examples

Worked examples using Denavit–Hartenberg parameters appear in educational materials from Massachusetts Institute of Technology, Stanford University, and publications by research groups at Carnegie Mellon University and University of California, Berkeley. Typical examples derive the forward kinematics for a 6-DOF industrial manipulator like models produced by KUKA and ABB Robotics, compute the Jacobian for singularity analysis performed in labs at ETH Zurich and Tokyo Institute of Technology, and compare DH results to PoE methods presented at conferences hosted by IEEE Robotics and Automation Society and IROS. Computational implementations are available in toolboxes distributed by MathWorks, open-source repositories on GitHub, and coursework at institutions such as University of Michigan and California Institute of Technology.

Category:Robotics