search

5355-ROStig / DAT295-av

Project in the course DAT295 - Autonomous and cooperative vehicular systems
0 stars 0 forks source link

readme

DAT295-av

A project in the course DAT295 - Autonomous and cooperative vehicular systems at Chalmers. We were 7 people in the group working with two different subprojects using wifibots:

  1. Cruise control/platooning using Lidar.
  2. Intersection arbitration with two bots in a four-way intersection.

Dependencies

How to run

I case any other students will work in the EG5355 lab in the future, here's how to run our projects.

Clone this repository and the forked roswifibot driver onto the wifibots:

# ask the TA/supervisor for passwords

# Assuming the configuration of the roof-mounted Linksys router hasn't changed
ssh wifitbot@192.168.1.102  # white antenna
git clone git@github.com:5355-ROStig/DAT295-av.git

# and the driver
cd DAT295-av
git clone git@github.com:5355-ROStig/roswifibot.git

# repeat for 192.168.1.103 (black antenna)
# and for 192.168.1.101 (lidar bot) if you want to run cruise control stuff

For each bot, go to the root of the repository (which is a catkin workspace) and run the catkin make tool.

cd DAT295-av
catkin_make

Remember to source the package files as described in the ROS documentation .

echo "/opt/ros/noetic/setup.bash" >> ~/.bashrc
# Might need to change to wherever you cloned it
echo "~/DAT295-av/devel/setup.bash" >> ~/.bashrc  
source ~/.bashrc

Braking and Platooning specific

Use Case 1

Place two robots one after another with a safe distance. The robots will be in the movement 3 sec and speed 0.2 (can be changed in the code).

For running experiment three files need to be activated. At front robot:

rosrun use-case-1.py front

At back robot:

rosrun use-case-1.py back

Start robots simultaneously (just on some computer that is contected to Rostig network):

python init_start.py

Finally, measure the distance between the robots after stop and compare with the initial one.

Use Case 2

To run this experiment it is enough to use only one robot with IR sensor and one obsticle in front (e.g. cardboard wall) which the robot will be going to. The robot is supposed to stop if the distance to the object is shorter than 50 cm. The speed used in the experiment was 0.2 (can be changed).

Start the robot:

rosrun use-case-2-py

The program will print the distance to the object.

Use Case 3

Place two robots one after another on a safe distance (70 cm was used). Back robot must be robot with a LiDAR-sensor.

Run firstly the back robot (it will adopt its distance to the front robot):

rosrun lidar_receiver.py

Run front robot:

rosrun use-case-3-front.py

Use Case 4

Place robots on the safe distance one after another.

Run first back robot:

rosrun use-case-4-back-new.py

Then run front robot:

rosrun use-case-4-front.py

Intersection arbitation specific

First start GulliView on the roof mounted computer:

ssh -X 192.168.1.122
# Go to the directory of this file
find . -iname "startCamerasBroadcast.sh"
# Run it
./startCamerasBroadcast.sh 2121

Then place two bots (only tested on the non lidar ones) on two different roads marked with blue tape. If this tape is gone you can look at the images in the mapdata node readme to at least make sense of the code.

Put AprilTags on the wifibot plywood mounts, then start the intersection arbitration on each bot.

# Open two terminals

# Run this on one bot (e.g. one with tag 6)
# For example, a scenario with priority road (huvudled)
roslaunch intercrossing_master wifibot.launch robot:=/ tag_id:=6 scenario='scenario1'

# Run this on the other bot (e.g. one with tag 4)
roslaunch intercrossing_master wifibot.launch robot:=/ tag_id:=4 scenario='scenario1'

Look at em go! Now you can log data with our time-to-collision package see the UDP messages being broadcasted with our monitor or use the visualization package to see the tags moving around in rviz.