All CAD files were printed using 5% rectilinear infill and no supports. A tutorial on how to construct the nipple can be found here: https://www.youtube.com/watch?v=TO4Haf6ZgnU
The following must be run within feeding_nipple_ws
which is the workspace to run the ATI Net F/T sensor.
Upload the FeedingNippleArduino.ino
to your Arduino. Calibrate the upper and lower bounds of the analogRead and its corresponding voltages to ensure accurate readings. Be sure the readings are printing properly on the serial monitor. Ensure the serial monitor is closed throughout the duration of data collection.
Turn on the ATI NET F/T sensor. Open your terminal. Remember to run source devel/setup.bash
everytime you open a new window in the root of your workspace to source it.
To initialize the core, type roscore
.
In a new window, launch the sensor topic by typing roslaunch netft_rdt_driver ft_sensor.launch ip_address:=192.168.1.6 bias:=True rate:=1000
.
To ensure the topic has been properly launched, type the command rostopic echo /ft_sensor/netft_data
. This will allow you to monitor the readings from the F/T sensor in real time.
Calibrate the F/T sensor by holding it in the air and typing the command rosservice call /ft_sensor/bias_cmd "cmd: 'bias'"
.
Run the bash script callibration.sh
in the feeding_nipple_ws to begin collecting data from the ATI NET F/T sensor.
Place the extract_forque.py
script in the same folder as your rosbag files. Run the extract_forque.py
script. This converts the rosbag into a CSV in its own individual folder.
Run this script with two CSVs files as the input - the voltage readings from the Arduino and the rosbag force readings. This script will combine the two csv files and calculate a calibration function in the form y = ae^(bx).
After finding the calibration function, input the a and b parameters into TranslateV2F.py
. Whenever performing future tests or in field tests, have FeedingNippleArduino.ino
running on the Arduino before running TranslateV2F.py
.