Peer-Reviewed Science
The technology powering DARERL’s Virtual Human Twins is built upon decades of foundational, peer-reviewed research in biomechanics, hyper-elastic simulation, and computational anatomy authored by our founding team.
Physics-based Animation
The highly cited, definitive textbook covering the full pipeline of physical simulation, from kinematics to collision detection. This work represents the holistic architectural knowledge required to build an enterprise-grade physics engine from the ground up.
View PublicationInteractive simulation of rigid body dynamics in computer graphics
A foundational state-of-the-art report on rigid body dynamics. This paper details the core mathematical solvers required to simulate real-time, interactive physical scenarios without computational breakdown.
Read Full PaperNon-smooth newton methods for deformable multi-body dynamics
A critical breakthrough for simulating human tissue. This research introduces methods to efficiently and stably compute interactions involving deformable, squishy bodies (like skin and muscle) interacting with complex, rigid structures (like devices).
Read Full PaperContact and friction simulation for computer graphics
A masterclass on formulating contact as a complementarity problem. This work dives deep into barrier functions and anisotropic friction, guaranteeing that the grip, slip, and tactile feedback generated by DARERL twins mirrors real-world physics.
Read Full PaperA hyper elasticity method for interactive virtual design of hearing aids
A direct application of DARERL’s core mission. This paper introduces a parallel method for general non-linear hyper-elasticity modeling, establishing how to interactively test hardware against the complex, squishy behavior of the human ear canal.
Read Full PaperDeep-learning-based segmentation of individual tooth and bone with periodontal ligament interface details for simulation purposes
Bridging the gap between AI and biomechanics. This methodology proves our capability to extract complex human biology from scans and build multi-layered twins that react to pressure just like living tissue interfaces.
Read Full PaperContact modeling from images using cut finite element solvers
This paper details advanced methodologies for simulating contact mechanics directly from medical imaging, bridging the gap between raw scanning data and actionable, physics-based simulations for product design.
Read Full PaperLocally orderless registration
An exploration into robust mathematical models that map and register varying biological data. This foundational research enables DARERL to confidently process diverse datasets of human anatomies to extract true statistical variance.
Read Full PaperA joint-constraint model for human joints using signed distance-fields
Crucial for creating Virtual Human Twins that move realistically. This model establishes how to mathematically constrain digital human joints so they mimic the exact range of motion and physical limits of a real human.
Read Full PaperAnalysis of deformation of the human ear and canal caused by mandibular movement
A vital study mapping how human anatomy physically changes during motion. By understanding how jaw movement deforms the ear canal, this research enables our twins to simulate dynamic biological strain rather than just static poses.
Read Full PaperOpen-Full-Jaw: An open-access dataset and pipeline for finite element models of human jaw
Demonstrates our capability to generate end-to-end, highly complex biological systems. By creating open pipelines for finite element modeling, we validate the robustness of the data structures fueling DARERL.
Read Full PaperDifferentiable depth for real2sim calibration of soft body simulations
This research validates how we tune our software to match reality. By comparing real-world soft-tissue deformation against simulated output, this framework allows the physics engine to automatically calibrate itself to perfectly mirror biological tissue.
Read Full PapergradSim: Differentiable simulation for system identification and visuomotor control
Pioneering work in differentiable simulation, empowering the physics engine to learn and adapt. This ensures that the digital interactions between virtual twins and digital prototypes can be optimized using modern AI and control systems.
Read Full PaperA systematic comparison between FEBio and PolyFEM for biomechanical systems
Rigorous benchmarking of biomechanical simulation engines. By actively testing and comparing computational limits, this research guarantees that the numerical methods backing DARERL are mathematically sound and industry-leading.
Read Full PaperCollision Constraints & Diffeomorphisms: Dealing with Physics in Deformable Image Registration
The absolute state of the art in medical image processing. This framework ensures that when extracting anatomy to build our Virtual Twins, tissue layers slide and deform accurately with zero physically impossible self-intersections.
Read Full Paper