# Data Collection

  1. Follow the steps here to initialize the server for battery power.
  1. Connnect to the server via SSH over the ethernet cable in the front passenger seat and initialize the docker:

    # ssh connection to the vehicle
    ssh autodrive@192.168.1.1
    
    # start the docker
    cd docker/
    ./run.sh

    You should then see a modified prompt for the docker: [DOCKER]

  2. Start a tmux session inside the docker container. The default keybinds are configured. For quick reference:

    Keybinds
    CTRL-B + " : Split window horizontally
    CTRL-B + % : Split window vertical
    CTRL-B + C : Create a new window
    CTRL-B + N : Move to next window
    CTRL-B + index : Move to window at index, ex: CTRL-B + 0 will move to the first created window
  3. Use roslaunch to launch the system for data collection in one window/pane:

    roslaunch bolt_launch data_collection.launch
  4. Using the topic analyzer TUI from a separate tmux window/pane:

    1. Load the data collection configuration file
    2. Ensure that all topics listed are publishing at the target rate (besides CAN topics and Piksi, those are published when outside/moving)
    3. If you see a topic struggling, or want to verify the topic data, focus that topic
  5. If you have ensured the integrity of topics, move the car outside.

  6. Start up the data collection script in a separate tmux window/pane:

     ./tools/data_collection.sh bag_name
    • bag_name should be descriptive relative to whats happening with the weather. Examples:
      1. campus_snow for snowy weather
      2. campus_rain for rainy weather
      3. campus_sunny for sunny weather
  7. Drive the planned route.

Route for Data Collection
Data Collection Route
  1. Park the vehicle in the marked area, and DO NOT TURN OFF THE CAR.
  1. Plug the vehicle into lab ethernet, and rsync the bags with our campus storage:
    ssh -L 9922:davinci.egr.msu.edu:22 {username}@compute01.egr.msu.edu
    rsync --progress -av -e "ssh -p 9922" /mnt/internal-datasets/{bag_name}*.bag \
      {username}@localhost:/egr/research-canvas/raw_datasets/campus/