Kinematic
Environment.
foundational multi-axis articulation kernel. iacon kinematics provides a deterministic bridge between semantic target states and physical motor actuation for high-dof organisms.
high-frequency state resolution
sub-micron error correction
parallel state vector ingestion
Algorithm_Kernel
Target_Vector_Stream
Live_Joint_State
| MORPHOLOGY_ID | DOF_COUNT | SOLVER_STATE | MEAN_TARGET_ERROR | SOLVER_THROUGHPUT |
|---|---|---|---|---|
M_01Manipulator | 6_DOF | OPTIMIZED | < 0.01mm | 4KHz |
M_02Quadruped | 12_DOF | OPTIMIZED | < 0.05mm | 2KHz |
M_03Humanoid | 28_DOF | OPTIMIZED | < 0.12mm | 1KHz |
M_04Dexterous_Hand | 22_DOF | OPTIMIZED | < 0.02mm | 2KHz |
Iterative CCD
highly optimized cyclic coordinate descent implementation. calculates n-degree of freedom poses without matrix inversions, guaranteeing convergence in < 1ms.
GJK / EPA Mesh
continuous collision detection (ccd) utilizing high-fidelity mesh intersection. ensures zero penetration even at hardware-native high frequencies.
URDF Ingestion
native parser for industry-standard urdf and mjcf files. automatic generation of joint hierarchies, mass properties, and rotational constraints.
Control Pipeline
direct integration with existing reinforcement learning pipelines. iacon kinematics is exposed via ultra-low latency python bindings, supporting batched gpu solving for thousands of concurrent environments.
import iacon.kinematics as ik# 1. initialize morphologysolver = ik.Solver(config="go2.urdf",solver_type="ccd_batch")# 2. define target posestargets = torch.tensor([[1.2, 0.4, -0.8]])# 3. resolve joint states (60fps)state = solver.resolve(targets)print(state.joint_angles)
Simulate Reality.
deploy the iacon kinematic backbone to your cluster. build high-fidelity motion policies for any morphology at scale.