Data Collection

Collect images and odometry from the robot

To develop your solution to this mapping exercise, you will collect testing data from the maze. You need to create another script to save images and odometry from the robot as you move it through the maze via the server-teleop.py script.

Create a new python script (data_logs.py) that logs the robots odometry (d, theta) in a text file and saves images to your laptop.

1. Write code to calculate the robots displacement (d) and angle (theta) from the wheel encoder ticks.
   • Store the change in encoder ticks from the last data point
   • Convert encoder ticks into a distance travelled per wheel
   • Calculate and store the robot displacement and angle (Hint: see equations below).
2. Write code to log data points that consists of:
   • a camera image saved to disk (see cv2.imwrite()).
   • a line in a text file in format: img_name d theta (Hint: text files in python).
   • Note: you should control the frequency new data points are recorded at (e.g only after moving every 3cm or 3 degrees as calculated in step 1)

Run this new script as you guide your robot through the maze (run server-teleop.py in a parallel terminal window) and collect your data!

Before moving on, verify that the images have been saved correctly and the odometry recordings are reasonable.

Odometry Equations

To calculate robot odometry from wheel distances:

To find change in distance, average the distance travelled by each wheel.

distance = (distance_left + distance_right) / 2.0

To find change in bearing, divide the difference between each wheel distance by the robot baseline (distance between wheels on the robot) and convert this angle to degrees.

theta = (distance_left - distance_right) / baseline  * 180/pi