Hardware Components

August 24, 2026 ยท View on GitHub

This file lists the hardware that matches the current active implementation.

Main Robot Computer

  • Raspberry Pi 5
  • Ubuntu / ROS 2 runtime
  • Docker-based robot stack
  • SSD storage for robot stack and logs

Motor Control

Current active motor path:

Raspberry Pi 5 -> GPIO -> DRV8833 -> motors

Active components:

  • Raspberry Pi 5 GPIO
  • DRV8833 motor driver
  • left and right DC motors
  • separate motor power supply
  • common ground between Raspberry Pi, motor driver and motor supply

Current bridge configuration:

BRIDGE_MODE=rpi_direct
DRIVER=drv8833
ENCODERS_ENABLED=0

Robot pose for navigation comes from LiDAR, SLAM, AMCL/Nav2 and available ROS pose topics.

Perception and Navigation Sensors

Active sensors:

  • RPLidar A1 for /scan, SLAM and Nav2
  • Intel RealSense camera for RGB/depth/camera_info/IMU topics
  • compressed image topics for Jetson inference transport

AI / Anomaly Detection Hardware

Active anomaly detection hardware:

  • Jetson Orin

Jetson responsibilities:

  • subscribe to selected Raspberry Pi topics through rosbridge
  • run YOLO inference
  • save original anomaly images, annotated images, map snapshots and JSONL event logs locally
  • publish anomaly visualization topics back through rosbridge

Visualization / Networking

  • rosbridge_server on Raspberry Pi, usually port 9090
  • foxglove_bridge on Raspberry Pi, usually port 8765
  • Foxglove on laptop or Jetson connects to Raspberry Pi foxglove_bridge
  • Jetson communicates with Raspberry Pi through rosbridge WebSocket