ROS 2 · Autonomous mobile robotics

Autonomous Patrol Robot Using ROS 2 + Nav2.

An autonomous mobile robot system using a TortoiseBot Pro Max for indoor patrol, built with ROS 2 Humble, Nav2, AMCL localization, occupancy-grid mapping, dynamic replanning, and custom mission-level autonomy.

Platform TortoiseBot Pro Max
Environment Gazebo simulation
Navigation Nav2 · AMCL
Mission Logic Custom Python Patrol Manager
01 / LOCALIZE
AMCL

Estimate the robot pose using LiDAR and an occupancy map.

02 / PLAN
Nav2

Generate paths to each patrol waypoint.

03 / CONTROL
Patrol Manager

Dispatch goals and monitor navigation status.

04 / RECOVER
Recovery Logic

Retry, clear costmaps, or skip unreachable waypoints.

01 / Problem

A patrol robot needs more than one-shot navigation.

A patrol robot must localize itself, visit multiple waypoints, recover from failures, and adapt when obstacles appear. Simply sending one navigation goal is not enough for continuous autonomous patrol.

02 / Solution

A complete navigation and mission-management stack.

I built a ROS 2 Humble simulation using Gazebo, Nav2, AMCL, LiDAR, odometry, and occupancy-grid maps.

I added a custom Python Patrol Manager node that loads patrol waypoints, sends NavigateToPose goals, monitors goal status, retries failed goals, clears costmaps, skips unreachable waypoints, and continuously loops through the patrol route.

Localization AMCL
Mapping Occupancy-grid map
Mission Interface NavigateToPose
Mission Control Python Patrol Manager

03 / Mission-Level Autonomy

Navigation behavior beyond the default Nav2 demo.

Load

Read a predefined patrol route containing multiple waypoints.

Dispatch

Send each waypoint as a NavigateToPose action goal.

Monitor

Track navigation feedback and the final status of each goal.

Recover

Retry failed goals, clear costmaps, and skip waypoints that remain unreachable.

Repeat

Continuously loop through the complete patrol route.

04 / Impact

Continuous and failure-aware autonomous patrol.

The system demonstrated autonomous goal execution, runtime obstacle avoidance, dynamic replanning, and mission-level patrol behavior beyond basic Nav2 navigation.

System Result The robot could execute a multi-waypoint patrol while responding to navigation failures and obstacles during operation.

05 / Simulation Videos

Autonomous patrol in simulation.

These simulation runs demonstrate multi-waypoint navigation, autonomous patrol execution, obstacle avoidance, and runtime replanning.

Multi-waypoint autonomous patrol

Obstacle avoidance and dynamic replanning

06 / Technologies

Navigation, simulation, and mission control.

ROS 2 Humble Nav2 AMCL Gazebo RViz Python LiDAR Odometry Costmaps YAML
Back to all projects ↗