Data and Simulation Infrastructure to Accelerate Intelligent Port Development
A simulation and evaluation baseline for port automation.
KYNPORT provides SimReady assets and procedural variant generation for port and logistics simulation. We deliver L2 asset packages ready to simulate and L3 data packages ready for training.
Pursuing domain standards to build compliant data assets.
Asset model joints are validated on real equipment, with joint topology, limits, mass-inertia and attachment points ready out of the box.
A deterministic, seed-driven generation pipeline produces compliant variants that are reproducible and traceable. Customers gain generation capability instead of purchasing model files one by one.
Across assets, generation, data and reconstruction, four product lines apply distinct frontier technologies to solve the challenges of building custom model assets.
Standardized assets for port and logistics simulation. Containers, RTG cranes and terminal tractors are generated to real industry standards with geometry, physics and semantics included. Run them in Isaac Sim or Gazebo, and take your assets with you—no platform lock-in.
Deterministic, seed-driven procedural generation creates liveries, size, wear, bulging and door-angle variants within the licensed scope. Choose from 12 shipping-line palettes, unbranded containers or the fictional KYNPORT brand. Every result is reproducible and traceable.
Build on L2 assets with domain-randomized variant sets, Replicator configurations and self-service generation scripts. Semantic labels carry container numbers and license plates, while instance segmentation aligns directly with the asset manifest—the foundation for trainable container OCR and smart tally systems.
For mines, construction sites and aging terminals without complete CAD: drone and LiDAR field capture turns point clouds into SimReady assets, delivered in tiers based on mechanism complexity. De-identified outputs strengthen the standard asset library while each customer project expands available coverage.
Procedural generation, multi-platform validation and correct implementation of domain standards form the technical foundation of KYNPORT SimReady assets.
Procedural modeling follows equipment specifications and real industry conventions, aligning corrugation pitch, corner-fitting apertures and owner liveries while keeping geometry and textures in phase.
Specification and seed inputs drive deterministic generation of variants and operating conditions across a randomization space that manual modeling cannot exhaust.
Joints, limits and operating conditions are replayed in Isaac Sim, Gazebo and other simulation platforms. Assets ship only after quantified checks pass, with reproducible results.
Specifications and seeds generate geometry and six-face textures with industry rules built in
Joint topology, mass, inertia and multiple sensors are produced together
Liveries, wear, weather, lighting and operating sequences are combined automatically
Ground-truth boxes, segmentation and instance IDs for containers and plates are generated automatically
Synthetic-data pretraining is followed by fine-tuning with a small set of real data
Replay tests in Isaac Sim, Gazebo and other platforms verify quantified criteria before delivery
Reproduce yard operations in Isaac Sim or Gazebo using five RTG systems—hoist, rotation, 40↔20 telescoping, gantry travel and trolley traverse—for solution validation and retrofit planning.
An automated evaluation standard built around format loading, joint travel, limits, static sag, attachment points, semantic labels and reproducibility.
Semantic labels include container numbers and license plates, with randomized weather, lighting and wear to support container OCR, smart tally and terminal tractor recognition.
Integrators can bring operational L2 asset packages into solution demos and digital twin deliveries. Assets remain portable and independent of any single platform.
Combine terminal tractors and containers with deck heights and container positions calibrated to real equipment for gate, yard and intermodal scenarios.
Drone and LiDAR field capture turns point clouds into SimReady assets, bringing aging terminals, mines and construction sites into simulation.
Decoupled visual, physics and semantic layers connect to mainstream simulation engines and training frameworks.
Hoist, rotation, 40↔20 spreader telescoping, gantry travel and trolley traverse have all been tested. Limits and control interfaces for 13 joints ship with the asset.
The first sample set includes containers, RTG cranes and terminal tractors, with import support for Isaac Sim and Gazebo. Details are subject to official announcements.
The real threshold from manufacturer CAD to SimReady assets lies in joint topology, mass and inertia, collision approximations, and contracts for units and orientation.
We are looking for asset engineers, platform engineers and business partners. If you understand ports and simulation—and want to turn simulation-ready into assets that can be verified—we would like to hear from you.
Apply at hr@kynport.comWhether you need an L2 asset package, L3 data package, 3D reconstruction or reverse engineering, share your requirements and we will respond within one business day.