The Autonomous Control Unit (ACU) is a high-performance edge computing device designed by YASKAWA to extend the capabilities of the YNX1000 robot controller. It serves as a platform for running advanced applications that require significant computational power, such as artificial intelligence and complex image processing. Its primary purpose is to provide a standardized environment where users can deploy custom applications while leveraging pre-built services for robot motion and perception.
Edge Computing:
The ACU executes compute-intensive tasks locally at the "edge" of the robot network to ensure low-latency performance.
Application Hosting:
The ACU runs user-developed applications packaged as Docker containers, allowing for a modular and portable software architecture.
Standard Services:
The ACU provides several built-in services for descrete robot control and advanced processing. Users can interact with these services directly with provided interfaces and skills. Additionally, developers can take full advantage of these services via their associated gRPC APIs. More information about these services can be found here.
The ACU is a compact industrial computer (NVIDIA Jetson Orin NX 16Gb) with the following key hardware features:
| Item | Specification |
|---|---|
| Interfaces | 4x Gigabit Ethernet (RJ-45), 3x USB 3.2, 1x USB 2.0 |
| Storage | 256 GB internal storage |
| Memory | 16 GB shared (RAM & VRAM) |
| CPU | 8-core ARM 64-bit |
| GPU | NVIDIA Ampere (1024 CUDA cores & 32 tensor cores) |
| Power | 24 VDC (±10%), 1.9A |
The ACU’s network architecture is designed to manage high-speed data transfer between the robot controller, peripheral devices, and development PCs. It utilizes four dedicated Gigabit Ethernet (GbE) ports to ensure low-latency communication and data isolation.
| Port | Default IP Address | Usage |
|---|---|---|
| LAN1 | 192.168.10.253 | Used for primary communications with the Robot Control module. |
| LAN2 | 192.168.1.253 | Pendant and general external communications. |
| LAN3 | 192.168.11.253 | Unreserved |
| LAN4 | 192.168.12.253 | Intended for industrial GigE camera connections. |
Users can configure and manage the network settings through the ACU System Manager web interface:
IP Address Configuration:
The interface allows for manual IP address entry or the use of DHCP for automatic assignment.
Interface Control:
Individual network interfaces can be enabled or disabled as needed.
Virtual Network Settings:
The System Manager also provides options to configure virtual networks, which is useful for managing containerized application traffic within the Docker environment.
There are two primary ways the ACU is utilized in a robot system:
Robot Controller Configuration:
The ACU is embedded directly into the YNX1000 robot controller. In this setup, it is internally wired for power and communication with the robot's main control module.
YNX DevKit Configuration:
This is a standalone development kit used for application development and verification. It allows developers to test their software in a virtual environment using the YNX Robot Simulator without needing a physical robot manipulator.
Building user applications for the ACU is made easier with the provided ACU SDK. This development kit allows developers to create custom services and skills. These can be understood as modular software components that perform specific tasks like object recognition or trajectory calculation. These skills can then be used by larger ACU applications or otherwise empower end-user functionality directly from standard INFORM. This enables the robot to perform highly specialized autonomous functions that are not available in the standard controller software.
| ⬅️ Previous | Next ➡️ | |
|---|---|---|
| What makes NEXT different? | What is Docker? |