Dipping apples in honey is a Jewish tradition on Rosh Hashanah, the Jewish new year, to symbolize a sweet new year. In this tutorial, you will program a robotic arm to dip apples in honey - no more sticky hands!

Watch the Instagram post to see a working demonstration.

Natalia with robot

Prerequisites

What You'll Learn

What You'll Need

What You'll Build

Add a new Viam machine

  1. In the Viam app, create a machine by typing in a name and clicking Add machine. add machine
  2. Click View setup instructions. setup instructions
  3. Install viam-server on the device that you want to use to communicate with and control your robot arm. For example, select the Mac platform if you want to use your Mac computer to control the arm. If you're using a different computer, select the platform appropriate for your setup. select platform
  4. Following the rest of the installation instructions on the setup page, install viam-server on your computer by entering the provided CLI command into your terminal prompt. The example shown here is on a Mac. install viam-server
  5. Download the machine cloud credentials file to your computer for viam-server to connect to app.viam.com.
  6. Run viam-server locally on your computer with the config you just downloaded. If your download in the previous step went to a different folder, modify the file path. 
    viam-server -config ~/Downloads/viam-shana-tobot-main.json
  7. The setup page will indicate when the machine is successfully connected. live machine

Configure your machine

Now that our machine is live, let's configure an arm component.

  1. Under the CONFIGURE tab for your live machine, click on the + symbol next to your machine in the left sidebar and select Component in the menu that opens. add component
  2. Choose arm as the type, and then choose your desired model. For example, if you're using a UR5e, choose the ur5e model from the list. select arm model
  3. Leave the default name, and then click Create. leave arm name
  4. In the newly created component panel, configure any attributes as needed. Refer to the documentation for the model for information about your arm's model. In the example shown here, we are using a ur5e arm model, inputting a required host field with the arm's IP address, and moving at a speed of 20 degrees per second. configure attributes
  5. Save your machine's configuration.

Control your machine from the Viam app

  1. Under the CONTROL tab of your live machine, find the corresponding panel for your configured arm. control with Viam app
  2. Experiment with getting joint positions to fetch data about the robotic arm's current position, and moving joint positions.

Control your machine with code

Now that you know how to work with your robotic arm using the Viam app, learn how to program your arm using code.

  1. Under the CONNECT tab of your live machine, select the Python SDK we'll use to access the arm. Follow the instructions to install the Python SDK if you haven't already.
    mkdir shana-tobot && cd shana-tobot
python3 -m venv .venv
source .venv/bin/activate
pip3 install viam-sdk
  2. Toggle on Include API key to see your machine credentials in the boilerplate code below. Make a note of your API-KEY, API-KEY-ID, and MACHINE-ADDRESS-URI for the next steps. show API key
  3. From the command line of your terminal window, create a new file to contain your script.
    touch script.py
  4. To program the movement of the robotic arm, copy and paste the following Python code into script.py using your preferred code editor. Be sure to replace the placeholders for <API-KEY>, <API-KEY-ID>, and <MACHINE-ADDRESS-URI> with your own machine credentials from the previous steps. 
    import asyncio

from viam.robot.client import RobotClient
from viam.components.arm import Arm
from viam.proto.component.arm import JointPositions
from viam.proto.common import Pose

async def connect():
  opts = RobotClient.Options.with_api_key(
      api_key='<API-KEY>',
      api_key_id='<API-KEY-ID>'
  )
  return await RobotClient.at_address('<MACHINE-ADDRESS-URI>', opts)

async def main():
  machine = await connect()

  print('Resources:')
  print(machine.resource_names)

  # move arm to exact end position then get coordinates:
  # arm
  arm = Arm.from_robot(machine, "arm-1")
  arm_return_value = await arm.get_end_position()
  print(f"arm get_end_position return value: {arm_return_value}")

  # hard coded end position - grabbed from terminal output
  x = -117.89177530820058
  y = 810.53648219876914
  z = -86.077008874770044
  o_x = -0.15705296827522847
  o_y = 0.24567021047557136
  o_z = -0.95654613732994043

  cmd_arm_pose = Pose(x=x, y=y, z=z, o_x=o_x, o_y=o_y, o_z=o_z)

  #Move to bowl
  await arm.move_to_position(pose=cmd_arm_pose)

  #adjust to go a little deeper into the honey
  my_arm_joint_positions = await arm.get_joint_positions()
 newJointPos = JointPositions(
     values=[
         my_arm_joint_positions.values[0],
         my_arm_joint_positions.values[1] + 6,
         my_arm_joint_positions.values[2],
         my_arm_joint_positions.values[3],
         my_arm_joint_positions.values[4],
         my_arm_joint_positions.values[5],
     ]
 )
  print("dipped")

  await asyncio.sleep(2)

  #what are current joints?
  my_arm_joint_positions = await arm.get_joint_positions()
  print(f"myArm get_joint_positions return value: {my_arm_joint_positions}")

  #lift joint 1 and joint 3, drop joint 2
 newJointPos = JointPositions(
     values=[
         my_arm_joint_positions.values[0],
         my_arm_joint_positions.values[1] - 40,
         my_arm_joint_positions.values[2] + 25,
         my_arm_joint_positions.values[3] - 30,
         my_arm_joint_positions.values[4],
         my_arm_joint_positions.values[5],
     ]
 )

  #move
  await arm.move_to_joint_positions(positions=newJointPos)
  print("lifted")

  # Don't forget to close the machine when you're done!
  await machine.close()

if __name__ == '__main__':
  asyncio.run(main())
  5. Save your changes, and then run the script from your command line.
    python3 script.py

Customize the robotic arm

In addition to continuing to refine the code, you can also refine the robotic arm. In our example, to dip a sliced apple in a bowl of honey, it was helpful to repurpose the gripper by adding a skewer (in this case a chopstick). For your own use case, a different utensil with different properties and form might be more helpful.

chopstick

The Viam arm component library has several methods to simplify accessing and working with robotic arms. Learn how to work with joint positions and poses to quickly adjust one or more elements of position AND orientation simultaneously, by modifying other elements of the original arm pose.

What You Learned

Related projects