Summer 2017

Tanya's Goals:

  1. Set up IMU (install firmware on board, install software libraries/tools on Jetson, and calibrate.)
  2. Develop ROS application to estimate and output odometry using only IMU data.
  3. Add IMU odometry to enable the ROS amcl package (adaptive Monte Carlo localization algorithm).
  4. Obstacle detection using ZED camera only.
  5. Odometry estimation using ZED camera only.
  6. Improved odometry package using both IMU and ZED data.
  7. {there will be more...}

This week's tasks:

Goals 1-2, 4

  • Pursue the plan of action discussed here.
      • Look at the firmware code for the SEN-14001 and the Arduino libraries (all in the SparkFun repo). Find out/implement the changes are needed in programming from the Arduino side in order to obtain outputs/message types that are of the same form as the ROS message type for odometry*.
      • * also please tell me what this message type is. Is there only one used? Are there different variations? What was used in the original razor_imu_9dof ROS package?
      • If the prior step works, look at the teensyduino firmware code for an example of using rosserial to communicate between a serial device and ROS. Then add this to the Arduino code for the IMU. So we will have two serial devices overall: IMU and TeensyDuino.
  • If previous step works out, load the firmware file on the actual hardware. Test out the code you modified from the previous step to get odometry messages on the ROS side.


  • Survey and compare existing algorithms for obstacle detection using stereo cameras

Completed tasks:

  • Write ROS code for odometry estimation using IMU data.
  • Find and use rosbag data to calibrate/test this code.

Nicole's Goals:

  1. Adjust chassis design and figure out alternative material/production method.
  2. Assemble TeensyDuino breakout board for the second car.
  3. Work with Tanya to get the new IMU part functioning.