Autonomous floor plan mapper with Jetbot AI

Use the nvidia jetbot kit to construct a floor plan autonomously.

Fork of https://github.com/dusty-nv/jetbot_ros

1. Initial Setup

Startup the jetbot and do the following basic setup:

Start the JetBot ROS2 Foxy container

git clone https://github.com/ThomasR155/jetbot_ros
cd jetbot_ros
docker/run.sh

Run JetBot

Start the jetbot motors and camera:

ros2 launch jetbot_ros jetbot_nvidia.launch.py

Then to run the following commands, launch a new terminal session into the container:

sudo docker exec -it jetbot_ros /bin/bash

2. Navigation Model Setup

Test Teleop

ros2 launch jetbot_ros teleop_keyboard.launch.py

The keyboard controls are as follows:

w/x:  increase/decrease linear velocity
a/d:  increase/decrease angular velocity

space key, s:  force stop

Press Ctrl+C to quit.

Train Navigation Model

Run this from inside the container, substituting the path of the dataset that you collected (by default, it will be in a timestamped folder under /workspace/src/jetbot_ros/data/datasets/)

cd /workspace/src/jetbot_ros/jetbot_ros/dnn
python3 train.py --data /workspace/src/jetbot_ros/data/datasets/20211018-160950/

3. SLAM Setup

Start Lidar

Install Rplidar Driver for ROS2 (from source)

Make sure you have a suitable Lidar connected and mounted to your jetbot via USB just like that:

Then start the Lidar driver by running:

ros2 run rplidar_ros rplidarNode

Calculate Odometry from Lidar

Install rf2o_laser_odometry (from source)

Odometry will be started automatically, if you want to manually start it run:

ros2 launch rf2o_laser_odometry rf2o_laser_odometry.launch.py

4. Autonomous Mapping

Start SLAM

Install slam_toolbox

Place the robot in your room, make sure it's unblocked.

Run slam + odometry with the correct settings:

ros2 launch jetbot_ros jetbot_custom_slam.launch.py

Run Navigation Model

Use your custom trained model to navigate the room autonomously. Substitute the path to your model below:

ros2 launch jetbot_ros nav_model.launch.py model:=/workspace/src/jetbot_ros/data/models/202106282129/model_best.pth

Create Floor plan

create the floor plan by letting the robot navigate through the room autonomously:

Full Demo Video on Youtube

Save finished map

ros2 service call /slam_toolbox/save_map slam_toolbox/srv/SaveMap "name: data:'map.pgm'"

5. Postprocessing

Load Map into Postprocessing.ipynb

Preprocess the map with rotation, inverse, thresholding etc.

Apply Blob detection inside the room to detect obstacles:

Apply Edge detection to detect most outer edge and calculate its area (on inverse image):

Construct final floor plan including outer edge, obstacles and calculated area:

Acknowledgements

We would like to express our gratitude to Prof. Dr. Patrick Glauner for giving very interesting lectures and providing the hardware for this project.

Project team: Thomas Riedl, Leon Madest, Darwin Sucuzhañay and Ankit Singh Rawat

LinkedIn Post