Arm Cam

The new and improved method for working with the PX-100 arm in the Autonomous Robotics Lab at Brandeis University

For help

Benjamin Blinder, [email protected]

Summary

The PX-100 Arm was a nightmare to use. This project aims to simplify and standardize the useage of the arm. It assumes that you are still using the large rig setup with a webcam in the green holder, and the arm screwed onto the base. It also assumes you are using a computer that has been properly set up with Ubuntu 20.04.6 and both the arm and camera are connected to the laptop

How To Use

Setup

Make sure you have the arm placed correctly on the base of the rig, with a camera placed in the green holder above it. This code might work in other settings, but I would insist upon using the rig I built for it.

If you are in the lab, make sure to use either of the laptops that are marked with pink tape and say "Works with the arm." If you use on of these arms, skip the "Running the Code" Section.

If you want to setup a new laptop for use with the arm, here is the a brief overview of the arm setup. It is worth your time to follow the setup instructions for Ubuntu here and make sure you use Ubuntu 20.04.6, as well as the instructions here for setting up the Interbotix Workspace.

Warning

I would highly recommend using one of the already working laptops. Some laptops simply do not have compatible hardware with the PX-100 arm and therefore will not work, and you will not know this until you've already spent hours setting up the laptop. That being said, the issue with hardware compatibility can be fixed by replacing the U2D2 Controller in the arm with a new one from Trossen Robotics.

1. Install Ubuntu 20.04.6 This is the only version that I know for sure works, but other versions of 20.04 might work.

2. Install Ros Noetic from this link

3. Install pip, these instructions seem fine but I cannot verify since I did not use it for my installation.

4. Using pip, run this command in terminal:

pip install tk

5. Install the Interbotix Workspace described here.

6. Clone the rosutils repo and follow the new VM instructions.

7. Correctly set up the Bashrc file with the template from the rosutils

8. Change Line 11 in control_arm.py to match the path to the interbotix workspace on your laptop

9. Make sure the arm is plugged into the laptop via one usb port in the laptop and the arm's power cable is plugged in.

10. Connect the camera to a second usb port on the laptop.

11. Clone this repo into the Catkin Workspace on your laptop, and run the catkin make command.

12. Open a terminal and type the following command to start the program:

roslaunch armcam arm_command.launch

Troubleshooting

• The most common issue is the arm not being detected correctly. This is when the RViz window opens but most of the arm is colored bright white and refuses to accept commands. To fix this issue, ensure that your interbotix workspace is correctly setup and the path is correctly set in your control_arm.py file.

  If the arm is still not being recognized after that, you will need to install the Dynamixel software and ensure that it finds the arm and all the motors correctly.

  If that fails, then your computer is most likely incompatible with the arm.

• If the camera is not being picked up, make sure your bashrc file is set up consistently with the rosutils template.

• If you want to use a standard robot along with the arm, then you will need to make a link between the bashrc $path and the bru.py file from rosutils. You might also need to ensure that tailscale is working and connects to the lab network correctly.

Using the Arm

By now, you should have the arm set up and the software capable of running

The useage of the arm is farily simple. The black tape on the base shows where an object can be reached from.

Run the program using te following command in the terminal:

roslaunch armcam arm_command.launch

The arm should move to a start location, and a Tkinter window should open (It does not have an icon in the sidebar, but it is labeled "Control Panel")

The controls on the Tkinter window are fairly easy to use. The 3 entry widgets at the top are x,y, and z coordinates for the arm, which are sent with the Send Coords button.

The start button will begin the automatic pickup of an object.

The Sleep button sends the arm to the sleep position.

The Home button sends the arm to the home position.

The red Open and Close gripper buttons open and close the gripper on the arm.

The time scale changes how long it takes for commands to run, and the default is 1.

At the bottom, the Controls button pulls up the standard control window (Which is what starts open) and the Color Test window opens up a live color testing window.

In the live color testing window, the program pulls up a live view of the binary image that is detected after processing the image. This can be changed to include a wider or narrower range of colors by changing the color arrays.

Color arrays are set in RGB values, the upper array is the brighter value that can be seen, and the lower array is the darker value. These can be adjusted live, but to save any changes to the color detection, you must change the arrays saved on lines 13 and 14 of color_mask_test.py (Which is only used during the live testing).

This will not change the colors used in picking up an object automatically unless you also change the chosen color arrays in the arm_cam.py file, lines 17 to 21.

Tour of included files:

• arm_command.launch
  • This is the launch command to open the program.
• arm_cam.py
  • A rospy file that deals with the camera directly. This processes the image and sends the coordinates to the box_pickup.py file.
• box_pickup.py
  • This deals with the logic for picking up an object. It sends commands to the control_arm.py file.
• control_arm.py
  • This is the file that sends commands to the arm itself. You publish messages here and it deals with the Interbotix workspace to control the arm.
• arm_GUI.py
  • This is the GUI that helps to control the arm.
• color_mask_test.py
  • This file is only launched from the GUI when using the live color testing mode in the GUI.
• TKExample.py
  • This is the same as the standard GUI but can be launched without ROS installed so you can test the GUI to ensure it works before adding it to the real GUI.
• send_command.py
  • This is a backup. If you break the GUI, running this file allows you to have some basic control over the arm through a terminal.