ArtiSynth is a 3D modeling platform that supports the combined simulation of multibody and finite element models, together with contact and constraints. While targeted at biomechanical and biomedical applications, it can also be used for general purpose mechanical simulation. It is freely available under a two-clause BSD-style open source license.

State of the Art Functionality

ArtiSynth is a 3D biomechanical simulation platform, developed at UBC, directed toward modeling orofacial (OPAL) anatomy.

ArtiSynth has been used to develop a variety of biomechanical models, including upper airway and oral structures such as the jaw, hyoid, tongue, soft palate and pharyngeal wall; a muscle activated FEM model of the face; a combined multibody/FEM model of the foot; point-to-point muscle models of the arm and torso; and detailed FEM models of individual models including fiber fields and tendon sheets. It is the simulation platform for the OPAL and Parametric Human projects, and has also been used to create airway models for use in articulatory speech synthesis.

Powerful Capabilities


Shoulder Biomechanics

BIGLab’s research on the subject of shoulder biomechanics is done collaboratively with researchers from B.C, Saskatchewan, and Ontario. Work is done extensively on biomechanical modeling software such as ArtiSynth, and in vitro. Specific areas of research include Glenohumeral joint stability, mechanical tasks modeled by inverse simulations, and shoulder injury rehabilitation initiatives. The themes of biomechanical shoulder simulation are OpenSim and ArtiSynth.

The Shoulder Model in ArtiSynth
Building on work already done on other platforms, this shoulder model takes controls of ArtiSynth’s powerful capabilites for executing forward and inverse-dynamics simulations. The model utilizes state-of-the-art muscle wrapping techniques, accurate physical reconstruction, and observation test data to predict and model the movement of the glenohumeral joint, and the behaviour of the local musculoskeletal system.

Novel Techniques

The shoulder model in development incorporates a number of novel biomechanical simulation techniques. Among them is a glenohumeral joint forces constraint, which helps provide the intrinsically unstable joint with the control it has in the real-life biological system. It also features controller which help to accurately adjust the position of the bones around joint. With observational data, the scapula and clavicle can be adjusted to positions relative to the humerus which help provide the shoulder joint with strength and mobility.

Inverse Simulation

Some simulation are run to try to predict the muscle exertions required for certain movements. In the video to the right, the model is uses ArtiSynth’s inverse simulation capabilites to predict muscle activation levels required for the humerus to follow the blue wireframe precisely.

Modal Reduction

Modal reduction is the non-linear reduction of a model’s degrees of freedom to minimize the computation complexity of its simulation. For large or complicated simulations, this is novel technique to overcome the limitations computing power in the face of massively complicated simulations. The theme of this research is ArtiSynth.

Uncovering Modes

By applying stimulation to our model of the tongue, we can analyze the output from the simulations to uncover the primary ‘modes’ of movement available. Using these modes, reducing the model to something with fewer degrees of freedom is possible.

Reducing the Tongue

Despite greatly reducing the degrees of freedom in our reduced model of the tongue, the response to stimulus is approximately the same as in the unreduced model. This has important ramifications for the feasability of some large or complex model simulations.


Simulating the complex surfaces of biological models presents a unique challenge. Limits with strictly geometric skinning methods lead researchers to pursue physically-based skinning techniques. Physically based techniques avoid the problems with geometric skinning such as unrealistic bulging or pinching near joints, but only with far higher computational cost. Physically modeling all the points of a skin mesh drastically increases the degrees of freedom of a model.

A Novel Method