Runtime
LimbBot Core runs whole-body control on robot.
Six products combine runtime control, cloud adaptation, Isaac-based training, safety, planning, and developer integration.
Each product can stand alone in evaluation, but production programs usually combine them.
LimbBot Core runs whole-body control on robot.
Studio adapts policies to new dynamics and tasks.
Sim trains and validates policies with Isaac.
Guard enforces limits and recovery behaviors.
Planner composes long-horizon loco-manipulation.
API connects ROS 2, REST, telemetry, and deployment.
Start with Core for runtime control, then add Sim, Studio, Guard, Planner, and API as the program matures.
On-robot whole-body control runtime for balance, locomotion, and payload-aware motion at the edge.
Learn moreCloud fine-tuning and dynamics adaptation for new bodies, payloads, terrains, and tasks.
Learn moreGPU simulation and reinforcement-learning pipeline built on NVIDIA Isaac for contact-rich robot motion.
Learn moreSafety layer for fall avoidance, force limits, stop modes, and event telemetry around learned control.
Learn moreAgentic loco-manipulation motion planner that composes sub-skills for long-horizon mobile manipulation.
Learn moreDeveloper SDK and integration toolkit with ROS 2 node, REST controls, telemetry, and deployment utilities.
Learn moreSimulation and cloud tools create approved policies. Core runs them on the robot. Guard bounds them. Planner sends task goals. API ties it into your stack.
| Product | Category | Primary job |
|---|---|---|
| LimbBot Core | Runtime | On-robot whole-body control runtime for balance, locomotion, and payload-aware motion at the edge. |
| LimbBot Studio | Cloud | Cloud fine-tuning and dynamics adaptation for new bodies, payloads, terrains, and tasks. |
| LimbBot Sim | Simulation | GPU simulation and reinforcement-learning pipeline built on NVIDIA Isaac for contact-rich robot motion. |
| LimbBot Guard | Safety | Safety layer for fall avoidance, force limits, stop modes, and event telemetry around learned control. |
| LimbBot Planner | Planning | Agentic loco-manipulation motion planner that composes sub-skills for long-horizon mobile manipulation. |
| LimbBot API | Developer | Developer SDK and integration toolkit with ROS 2 node, REST controls, telemetry, and deployment utilities. |
Use LimbBot Sim to train and test policies.
Use Studio to tune dynamics and tasks.
Run Core with API integration on robot compute.
Guard monitors safety events and Planner composes tasks.
Product interfaces align around NVIDIA and ROS 2 workflows.
Training and validation environments.
Edge inference target.
Robot middleware integration.
Scene and digital twin workflows.
“The suite matters because runtime alone is not enough. We need simulation, safety, and integration gates to move toward fleet trials.”
A pilot can be narrow, then expand as the robot program matures.
Best for first hardware validation.
Best for adapting to new terrain and payloads.
Best for task workflows and fleet integration.
No. The right mix depends on program stage, but production deployments often combine several.
LimbBot Core runs the control runtime on robot edge compute.
LimbBot Guard handles fall avoidance, limits, and event telemetry.
LimbBot API provides ROS 2, REST, telemetry, and SDK tools.
Tell us where your robot program is today.
Request product guidance