Project Unicorn - intnav

Project Unicorn is a project for AMOD Fall 2018 (ETH Zürich) course that focuses on Intersection Navigation for duckiebots in Duckietown.

duckietown-intnav contains the code to implement the intersection navigation demo on a duckiebot. The general idea of the projects consists of a 4-step approach to efficiently navigate the 3 and 4-way intersections:

1- Starting at a red line at any intersection, the duckiebot estimates its initial pose.

2- Different paths for going left, right and straight are pre-computed and chosen depending on the desired intersection exit.

3- Constantly updating its pose estimate, it follows the path in a closed-loop manner with a PurePursuit controller.

4- Detects when the duckiebot has reached the intersection end (exit lane) and switches back to the duckietown lane follower.

The interested reader can find more details in the Project documentation.

Software architecture

The code is split into two sections:

• lib-intnav: contains the logic architecture and algorithms of the demo implementation. It contains the following nodes:

  • camera_config
  • kalman
  • planner
  • controller
  • imap (used for visualization purposes only)

• ros-intnav: contains all nodes that enable the communication between lib-intnav (algorithms for calculation) and ROS (communicates with the duckiebot). It contains the following nodes:

  • image_processing: Uses image_calibration.
  • april_activator
  • localization: Uses kalman.
  • controller: Uses controller, planner.
  • interface
  • tf_april_static: Uses tf (external library).
  • tf_cam_vehicle: Uses tf (external library).
  • visualization_imap: Uses imap.

Installation from source

We recommend to run our package natively:

$ cd lib-intnav
$ pip install -r requirements.txt --user
$ python setup.py develop --no-deps --user

To run the package using a virtual environment created by pipenv the following steps are necessary:

$ pipenv install
$ pipenv shell
$ cd lib-intnav
$ pip install -r requirements.txt
$ python setup.py develop --no-deps

Unit tests

Run this:

$ make -C lib-intnav tests-clean tests

The output is generated in the folder in lib-intnav/out-comptests/.