Back to case studies

Robotics

Autonomous Crawlspace Inspection Robot

Developed an autonomous crawlspace-inspection robot using ROS2, Nav2, SLAM Toolbox, frontier exploration, AI-based hazard detection, and remote iOS-based operation.

  • ROS2
  • Nav2
  • SLAM Toolbox
  • FastAPI
  • Raspberry Pi
  • Computer Vision
  • Frontier Exploration
  • iOS Application

Summary

Engineering context

Developed an autonomous crawlspace-inspection robot using ROS2, Nav2, SLAM Toolbox, frontier exploration, AI-based hazard detection, and remote iOS-based operation.

Category
Robotics
Year
Jan 2025 - Present
Status
Research and Prototype Development
Context
University of Regina MSc (Jan 2025 - Present)

My Role

Robotics and Autonomous Systems Engineer

Technical Stack

  • ROS2
  • Nav2
  • SLAM Toolbox
  • FastAPI
  • Raspberry Pi
  • Computer Vision
  • Frontier Exploration
  • iOS Application
  • Autonomous Robotics
  • Inspection Robotics
  • SLAM
  • Autonomous Navigation
  • Mobile Robotics
  • AI Inspection Systems
  • Robotics
  • iOS

System Architecture

  • ROS2 provided distributed robotic middleware architecture
  • Nav2 handled autonomous navigation and path planning
  • SLAM Toolbox generated real-time crawlspace mapping
  • Frontier exploration enabled autonomous environment exploration
  • AI perception systems detected mold growth and crawlspace hazards
  • FastAPI backend enabled communication between robot systems and mobile applications
  • iOS application provided remote monitoring and operation capabilities

Engineering Challenges

  • Autonomous navigation in confined crawlspace environments
  • Stable SLAM operation in low-feature environments
  • Developing reliable frontier exploration behavior
  • Integrating AI-based environmental inspection systems
  • Building real-time communication between robot and mobile applications

Hardware / Firmware / Software

Hardware

  • Autonomous mobile robot platform
  • Raspberry Pi onboard compute system
  • LiDAR sensing systems
  • Environmental sensing hardware
  • Vision-processing systems
  • Mobile robotics platform

Firmware

  • Embedded robotic control systems
  • Sensor integration firmware
  • Mobile robotics interface systems

Software

  • ROS2
  • Nav2
  • SLAM Toolbox
  • Frontier exploration algorithms
  • FastAPI backend
  • iOS application
  • AI hazard-detection systems
  • Environmental inspection systems

Sensors

  • LiDAR systems
  • Vision-based inspection systems
  • Humidity monitoring systems
  • Environmental monitoring sensors

Protocols

  • ROS2 DDS communication
  • HTTP API communication
  • Wi-Fi networking

Results / Outcomes

  • Successfully implemented ROS2 autonomous navigation stack
  • Demonstrated SLAM-based crawlspace mapping
  • Developed autonomous frontier exploration system
  • Integrated AI-based mold and animal detection
  • Developed remote iOS-based robot interaction platform
  • Demonstrated integrated mobile robotics inspection architecture

Gallery

Engineering Notes

Autonomous Navigation and SLAM

The robot used:

  • ROS2
  • Nav2
  • SLAM Toolbox
  • LiDAR sensing

to autonomously map and navigate confined crawlspace environments.

The system generated real-time occupancy maps and supported autonomous movement through unstructured environments.

Frontier Exploration

A custom frontier-exploration system was developed to enable autonomous exploration of unknown crawlspaces.

The exploration system:

  • identified unexplored map frontiers
  • generated autonomous navigation goals
  • expanded mapped environments automatically
  • supported fully autonomous inspection workflows

AI-Based Hazard Detection

Computer-vision systems were developed and trained to detect:

  • mold growth
  • crawlspace animals
  • environmental hazards

The AI perception pipeline enabled automated inspection and environmental analysis during autonomous exploration.

Mobile Application and Backend Integration

The robot communicated with an iOS application through a FastAPI backend running on the onboard Raspberry Pi system.

The architecture enabled:

  • remote robot monitoring
  • live telemetry
  • map visualization
  • remote interaction
  • mobile robot control

The iOS application interface was developed using AI-assisted software-development workflows.

Proof-of-Concept Demonstrations

Demonstrations included:

  • real-time SLAM visualization in RViz
  • autonomous frontier exploration
  • remote mobile interaction
  • autonomous inspection workflows

The system demonstrated integration of autonomy, perception, mapping, and mobile operation into a unified inspection robotics platform.