A browser-based robotics simulation suite featuring SLAM mapping with LiDAR, 6-DOF robotic arm kinematics, and multi-robot warehouse logistics. Built with Vanilla JS, Three.js, and HTML5 Canvas — no framework required.
Each engine is a fully self-contained simulation environment.
Simultaneous Localization and Mapping with 360° LiDAR raycasting, Bayesian occupancy grids, particle filter MCL, and autonomous frontier exploration. Three drive models, three pathfinding algorithms.
6-DOF robot arm with forward/inverse kinematics, LSPB trajectory planning, Teach Mode, and dynamic torque analysis per joint.
Multi-robot warehouse with 4 autonomous agents across 3 industrial scenarios — Warehouse, Assembly, and Mapping. Isometric 2.5D engine, battery management, collision avoidance, live analytics.
Every core algorithm is hand-coded in vanilla JavaScript.
TypedArray-optimized with wall proximity penalties and diagonal movement.
AStar.js + hive.html→Optimal shortest path without heuristic for guaranteed minimum cost.
Dijkstra.js→Reactive wall-following navigation for obstacle-dense environments.
BugAlgorithm.js→Autonomous frontier cell detection for complete unknown area mapping.
FrontierExplorer.js→Log-odds occupancy mapping with Bresenham ray voting for SLAM.
Mapper.js→Monte Carlo Localization with 50–500 configurable particles.
ParticleFilter.js→Scan-matching keyframe system for drift correction over long traversals.
LoopClosureDetector.js→2D raycasting with 30–360 rays, Gaussian noise, and ray visualization.
Lidar.js→DH parameter chain evaluation for 2–6 configurable joint configurations.
nexus.html→Numerical Jacobian IK solver for end-effector position targeting.
nexus.html→Linear Segment with Parabolic Blend for smooth joint trajectory planning.
nexus.html→Analytical dexterity index w = √(det(JJᵀ)) with Damped Least Squares IK singularity damping.
nexus.html→4D multi-agent path finding with dynamic corridor conflict resolution and priority yielding.
hive.html→Extended Kalman Filter state estimation with 1σ/2σ/3σ dynamic covariance uncertainty ellipses.
slam.html→EVERY PIXEL DRAWN WITH NATIVE BROWSER APIS
LiDAR mapping, Monte Carlo localization, frontier exploration, and three pathfinding algorithms.
LAUNCH SLAM6-DOF robot arm with forward/inverse kinematics, LSPB trajectory planning, and Teach Mode.
LAUNCH NEXUSMulti-robot warehouse logistics with isometric 2.5D rendering, battery sim, and task scheduling.
LAUNCH HIVEOPEN TO COLLABORATIONS, FEEDBACK, AND OPPORTUNITIES