Reference Design
Robotics Control System
A complete reference design for building production-ready robotics systems with advanced motion planning, sensor fusion, and real-time control for industrial and collaborative applications.
Introduction
What This Reference Design Is About
This robotics reference design provides a complete blueprint for developing production-ready robotic systems, from industrial manipulators to collaborative robots and autonomous mobile robots (AMRs). It demonstrates our proven approach to achieving precise motion control while ensuring safe human-robot interaction.
The design covers kinematic modeling, motion planning, sensor integration, and safety systems—enabling flexible automation solutions that adapt to changing production requirements.
Requirements
- Precision: ±0.01mm repeatability with advanced calibration
- Real-Time: 1kHz control loop with deterministic communication
- Safety: ISO 10218-1 and ISO/TS 15066 compliant for cobots
- Integration: ROS 2 compatible with industrial protocols
- Flexibility: Support for 3-DOF to 7-DOF configurations
Recommended Architecture
Distributed robotics architecture with separation of motion control, perception, and planning:
- • Motion Controller: Real-time servo control with EtherCAT master
- • Perception System: AI-powered 3D vision for object recognition
- • Planning Engine: Path planning with collision avoidance
- • Safety Controller: Independent safety monitoring (SIL3/PLe)
- • Application Layer: ROS 2 integration for high-level control
Essential Parts of This Design
- • Multi-axis servo control with trajectory interpolation
- • Forward/inverse kinematics for various robot configurations
- • Motion planning with MoveIt 2 integration
- • Force/torque sensing for compliant manipulation
- • 3D vision system with depth cameras
- • Safety-rated speed and separation monitoring
- • Teach pendant interface for programming
- • EtherCAT/PROFINET industrial connectivity
This reference design serves as a customizable foundation—adapt it to your specific robot type, payload, and application requirements.
Hardware
Recommended Platform: Arches + Joshua
Hybrid Platform: Arches (Jetson) for perception + Joshua (Sitara) for motion control
Why This Combination:
- • Arches: GPU acceleration for vision and AI inference
- • Joshua: Deterministic real-time for motion control
- • Clean separation of concerns for safety certification
- • Scalable to different robot configurations
Communication: EtherCAT or Ethernet/IP between platforms with sub-ms latency
Essential Components to Add
Components added to baseline platforms:
- • EtherCAT Master: Beckhoff ET1100 or equivalent for servo communication
- • Servo Drives: Multi-axis servo amplifiers with STO
- • Force/Torque Sensor: ATI Mini45 or OnRobot HEX
- • 3D Camera: Intel RealSense D455 or Photoneo
- • Safety PLC: Pilz PNOZ or Sick Flexi Soft
- • Teach Pendant: Custom HMI with E-stop integration
- • I/O Modules: Safety I/O for interlocks and sensors
Schematic Capture & PCB Layout
Design Tools: Altium Designer with industrial automation libraries
System Integration: Modular backplane design for flexibility
Key Considerations:
- • EtherCAT signal integrity for reliable communication
- • Safety relay integration with redundant paths
- • EMC design for industrial environments
- • Thermal management for continuous operation
Regulatory Requirements
Robot Safety:
- • ISO 10218-1/2 industrial robot safety
- • ISO/TS 15066 collaborative robot safety
- • IEC 62443 cybersecurity requirements
Functional Safety:
- • IEC 62061 SIL3 for safety functions
- • ISO 13849-1 PLe for safety-related controls
EMC: IEC 61000-6-2/4 for industrial immunity/emissions
RTOS/Operating System
Recommended: Linux + RTOS Hybrid
Ubuntu with real-time kernel for planning/perception + TI-RTOS for motion control
Arches (Jetson) - Linux:
- • Ubuntu 22.04 with PREEMPT_RT
- • ROS 2 Humble for robotics framework
- • NVIDIA Isaac for perception
Joshua (Sitara) - RTOS:
- • TI-RTOS for deterministic control
- • EtherCAT master stack
- • Direct servo loop execution
Why Both Are Needed
Linux Required For:
- • ROS 2 ecosystem and MoveIt 2 planning
- • Deep learning inference for vision
- • Complex motion planning algorithms
- • Cloud connectivity and analytics
RTOS Required For:
- • 1kHz servo control loops
- • Guaranteed latency for safety functions
- • EtherCAT cycle synchronization
Interprocessor Communication
Communication Architecture:
- • EtherCAT for motion commands (1kHz cycle)
- • Shared memory for high-bandwidth sensor data
- • UDP/TCP for configuration and diagnostics
ROS 2 Integration: ros2_control with hardware interface to motion controller
Kernel & Driver Configuration
Linux Kernel:
- • PREEMPT_RT for reduced latency
- • CPU isolation for ROS 2 executors
- • Memory locking for real-time threads
Custom Drivers:
- • EtherCAT master driver (SOEM/IGH)
- • Camera drivers with GStreamer pipeline
- • Force/torque sensor interface
- • Safety I/O driver
Middleware
Recommended Middleware: ROS 2
ROS 2 Humble - Industry-standard middleware for robotics with real-time capabilities.
Why ROS 2:
- • Extensive robotics ecosystem and community
- • MoveIt 2 for motion planning
- • Nav2 for mobile robot navigation
- • DDS for reliable real-time communication
- • Lifecycle management for production systems
Regulatory Considerations
Industrial Protocols:
- • EtherCAT for deterministic motion control
- • OPC UA for IT/OT integration
- • PROFINET for Siemens ecosystem
Safety Communication:
- • FSoE for safety over EtherCAT
- • Separate safety PLC for certified functions
Middleware Interfaces
Motion Interface:
- • ros2_control hardware interface
- • Joint trajectory actions
- • Cartesian path planning
- • Force/impedance control
Perception Interface:
- • Point cloud processing
- • Object detection and pose estimation
- • Occupancy mapping
- • Camera calibration
Application Interface:
- • Action servers for robot programs
- • Service calls for configuration
- • Diagnostic topics
- • Parameter server
Application
Application Architecture
ROS 2-based Embedded Linux Application with modular node architecture for flexibility.
Architecture Pattern: Component-based with lifecycle nodes for robust operation and clean startup/shutdown.
Design Decision: C++ for real-time nodes, Python for high-level orchestration and testing.
Core Application Modules
- • Motion Planner: MoveIt 2 with custom planners for constrained motion
- • Perception Stack: 3D vision with object detection and pose estimation
- • Trajectory Executor: Real-time interpolation with look-ahead
- • Force Controller: Admittance/impedance control for manipulation
- • Safety Monitor: Speed and separation monitoring
Development Tools
Essential Tools:
- • ROS 2 development environment
- • Gazebo for physics simulation
- • RViz 2 for visualization
- • PlotJuggler for data analysis
Testing Tools:
- • Robot simulation in Gazebo/Isaac Sim
- • ros2_test for integration testing
- • rosbag2 for data recording/playback
Diagnostics & Security
Log Management & Debugging:
- • ROS 2 logging with severity levels
- • Diagnostic aggregator for system health
- • rosbag2 for incident recording
Security:
- • ROS 2 SROS2 for encrypted communication
- • Access control for robot programs
- • Secure OTA updates via SWUpdate
Other Essential Supporting Apps
Cloud Integration
- • Fleet Management: Multi-robot coordination and monitoring
- • Digital Twin: Real-time simulation and visualization
- • Analytics: Cycle time optimization and OEE tracking
- • Remote Support: Secure remote access for troubleshooting
Desktop Applications
- • Robot Programming: Graphical program editor with simulation
- • Calibration Tool: Kinematic and camera calibration wizards
- • Offline Programming: CAD-to-path generation
- • Analysis Tool: Cycle time and motion analysis
Web-Based Apps
- • Teach Pendant: Web-based robot control interface
- • Monitoring Dashboard: Production status and KPIs
- • Program Library: Robot program management
- • Documentation: Interactive API and user guides
Our Experience
Products Delivered
Using this reference design, we have successfully delivered:
- Collaborative Robot: 6-DOF cobot with force sensing for assembly applications with PLe safety rating
- Warehouse AMR: Autonomous mobile robot with navigation and manipulation for logistics
- Welding Robot: 6-axis industrial robot with seam tracking and adaptive welding
Key Achievements
±0.01mm
Repeatability
PLe
Safety Rating
1kHz
Control Loop
7-DOF
Max Configuration
Our robotics reference design has been deployed in manufacturing, logistics, and research applications worldwide.