The course AERSP597 looks at the way humans interact with machines, specifically vehicles, using an approach founded in Cognitive Systems Engineering.

The participants in the course will also act as the humans in the human-machine system. The machine you will be using here is a simulation of an ice sailing craft or ice boat. This simulation was originally created for a course at TU Delft on distributed simulation, and is now adapted and expanded to run on Python. It is a vehicle for which no displays exist, and the current simulation is equipped with only a simple compass and a speed indicator. You can sail it, crash it, and we will upgrade the simulation to include other teams as well, so you will be able to crash the craft into each other and crash them against the shore.

Ultimately we will run a sailing race with these craft.

The basic simulation is created in two files. One is iceboat.py, this file provides the vehicle dynamics and loads the visual models. Your instructors will over time update this to make the competition more challenging. The second file is hud.py. The HUD currently shows a simple compass and prints the speed, and it is meant as a starter. The hud.py file is thus yours to modify, you should create the instruments, using the theory in the course, that help you sail a successful race. The design of these instruments must be motivated, and we will test design and theory in the races.

The simulation uses the widely available open-source programming language Python. The programming done in the course is limited, and you should be able to get by mostly by modifying the given examples.

You will need a number of software components to run the simulation. The software should run on major platforms (Windows, Apple Mac OSX, Linux), and packages are available on the conda ecosystem.

Steps to take for installation:

1. Install the Anaconda python distribution, download from the anaconda site and install. Use the Python 3 version, Python 2.7 will be obsoleted in January 2020.

2. When on Windows, open the Anaconda powershell prompt, to enter commands to configure your conda installation.

   On Linux or Linux or Mac OSX, you can simply open a terminal, and first enter the following command to activate the conda Python installation:

   conda activate
   

3. From the conda powershell, or your Linux/OSX shell with conda actated, enter:

   conda update conda
   conda update --all
   

4. Now add conda-forge, an additional software channel

   conda config --add channels conda-forge
   conda update --all
   

5. We need three pieces of sofware specifically for the simulation. git will be used to get the simulation source, Panda3d gives the visualization, and pyode with libode calculates the dynamics of the ice sailer. Because Panda3d is not in a conda package, we will use another program, pip, to install that. If you already have git, you do not need to install it through conda.

   conda install git
   conda install libode
   conda install pyode
   conda install pip
   conda install websockets
   pip install panda3d
   

   NOTE: libode and pyode were recently (Jan 2020) added to the conda-forge distribution.

6. Now get the source. I am hosting this on the gitlab of the Delft University of Technology.

   git clone https://gitlab.tudelft.nl/rvanpaassen/cse-vehicle-sailer.git
   

7. To start sailing, enter the newly created folder and run with python:

   cd cse-vehicle-sailer
   python iceboat.py
   

   If you have a file called iceboat.conf (see iceboat.conf.example), then the simulation will try to connect with a server. This server may initially run on your own machine, over the local network interface.

   With the server, you can add course elements to the scene (buoys and a referee boat), the wind can be modified, and the start location is now given by the server. Note that the objects can also be 'felt' in the simulation, if you crash into them you get stopped, and depending on the wind direction you might also get stuck.

   Likewise the shores of the simulation also affect the dynamics.

   You can start the server by running:

   python server.py
   

   The server reads the file server.conf for its configuration.

   You can also run these programs from the 'Spyder' editor. To run multiple simulations simultaneously, you need to re-configure the run mode in Spyder; select 'Run', 'Configuration per file' and under Console select 'Run in an external system terminal'.

   Note that we currently have a temporary database, the large green and red tubes are not visible in the simulated dynamics (you can sail through them), and point north and east; these are used for testing, and for getting a feel of the dynamics.

8. If there are updates to the simulation or support files, you can get these locally by entering:

   git pull
   

• Remove the testing x and y axis visualization

• Make the wind a little variable

• Add the markers for the race track

• Create proper interfaces (your work!)

PANDA3D 3D rendering framework.

Open Dynamics Engine physics library.

NumPy scientific computing with Python.

Python for beginners.

The Canvas page for this course.

Blender was used to create the 3d models.