Platform architecture
Wandelbots NOVA’s operating system provides a hardware-agnostic abstraction over robot controllers from manufacturers such as ABB, FANUC, KUKA, UR, and Yaskawa.
It’s part of a platform structured as four layers with open interfaces, so each layer can evolve independently and providers can be swapped without lock-in:
| Layer | Role |
|---|---|
| Application | End-user automation apps and operator interfaces |
| Execution | Hardware-agnostic robot control, motion planning, virtualization |
| Data and AI | Data pipelines, AI models, digital twins |
| Infrastructure | Compute, networking, storage |
The operating system’s job at the execution layer is to abstract over heterogeneous robot brands and present a single, consistent interface to everything above it.
This is delivered through the unified controller interface.
Unified controller interface
Wandelbots NOVA provides access to two types of controllers: physical and virtual. Physical controllers connect to robot hardware from manufacturers like ABB, FANUC, KUKA, UR, and Yaskawa. Virtual controllers simulate physical controllers in software.
Both are accessed through the NOVA controller API where you can individually configure them. Your application code works with both virtual and physical controllers. Only the controller identifier changes when switching between virtualization and production.
Development workflow
Develop with virtual controllers
Build and test application logic without hardware constraints or safety concerns. Use for development, testing, offline programming, and digital twins.
Clone physical configurations
Copy your physical controller configuration to a virtual controller to create an exact digital twin. Test configuration changes and application updates safely.
Deploy to physical controllers
Switch from virtual clone to physical robot by changing the controller identifier in your requests. Follow the robot setup guidelines to connect your physical robot hardware to NOVA.
Motion behavior
While you’re using the same endpoint for both controller types, their motion behavior differs.
Virtual controllers execute with idealized motion. Commands execute instantly, trajectories follow planned paths exactly, and round-trip latency is approximately 30ms.
Physical controllers exhibit real-world constraints. Motion depends on manufacturer-specific controller cycle times, mechanical inertia, and safety systems. Round-trip latency ranges from 100-300ms depending on the manufacturer.
Use the external joints stream to inject custom motion behavior from external simulators into virtual controllers.
Virtual controller endpoints
Virtual controllers mimic physical controllers. The unified controller interface provides access to core robot control functions. Virtual Controller endpoints replace teach pendant functionality like:
- Configuring tool center points (TCPs) and payloads
- Setting operation mode and emergency stop state
- Writing IOs
Virtual controllers also offer extended capabilities via the Virtual Controller Behaviour API where you can modify cycle time and inject custom motion behavior from external sources like simulators.
Virtualization
Virtualization in Wandelbots NOVA describes all processes and results connected to creating a virtual impression of a physical robot cell. This virtual impression is used to create robot programs and test them in a virtual environment before deploying them to a physical robot cell.
The 3D view in Robot Pad offers a visual representation of the virtual cell and the planned paths.
Virtualization is not the same as simulation. Virtualization is the in-platform digital representation of a controller.
Simulation is an external accurate virtual representation of a robot cell, used to test robot programs before deploying them to a physical cell. With Wandelbots NOVA, simulation is uniquely associated with using NVIDIA Isaac Sim.
Learn how to simulate with NVIDIA Isaac Sim and Wandelbots NOVA in the simulation introduction.