5. Run MAPLE and get pose data

MAPLE runs in a Docker container which makes it easy to deploy on different hardware and software environments. You can learn more about Docker here.

5.1. Starting and stopping

Navigate to the directory with your docker-compose.yml file in it (~/maple-config by default).

5.1.1. Start in background

docker-compose up -d

MAPLE starts in the background, attempts to restart if it fails, and will automatically start when the coprocessor is powered on.

5.1.2. Start in foreground

docker-compose up

MAPLE starts in the foreground, displays logs to the terminal, and can be stopped by pressing ctrl-c. Useful when configuring and calibrating cameras. Will not automatically restart if it fails or start when coprocessor is powered on.

5.1.3. Stop

docker-compose down

Stops MAPLE running in the background. Will disable MAPLE from starting when coprocessor is powered on.

5.2. Logging

A running log file of all MAPLE info, warning, and error messages is available in ~/maple-config/logs/maple_log.txt. If trajectory logging is enabled, a new logfile will be created in ~/maple-config/logs each time MAPLE is started.

5.3. Pose data on Network Tables

If a team number is set in the config.yml file, MAPLE will attempt to connect to the Network Tables server using the team number.

Pose data is available on the table MAPLE on topics position and orientation.

Topic	Description
position	Vector of robot pose x, y, and z in meters.
orientation	Vector of robot orientation roll, pitch, and yaw in degrees.
robot_relative_tagid	Vector of detected robot relative tag IDs. No duplicate tag IDs. The best tag detection (lowest error) across all cameras.
robot_relative_x	Vector of detected robot relative tag x distances, i.e. the distance from the robot center to the tag in the x axis. See Calibrate & configure your cameras for axis conventions.
robot_relative_y	Vector of detected robot relative tag y distances, i.e. the distance from the robot center to the tag in the y axis. See Calibrate & configure your cameras for axis conventions.
robot_relative_yaw	Vector of detected robot relative tag yaw angles, i.e. the rotation in degrees from robot forwards to tag forwards. 0 degrees means the robot is perfectly parallel to the tag. See Calibrate & configure your cameras for axis conventions.

5.3.1. Code examples

Access the position and orientation arrays from NetworkTables like any other array.

#include <networktables/DoubleArrayTopic.h>
#include <networktables/NetworkTablesInstance.h>

nt::DoubleArraySubscriber positionSub;
nt::DoubleArraySubscriber orientationSub;

void Robot::RobotInit() override {

  // Add to your RobotInit, Command, or Subsystem
  auto table = nt::NetworkTableInstance::GetDefault().GetTable("MAPLE");
  positionSub = table->GetDoubleArrayTopic("position").Subscribe({});
  orientationSub = table->GetDoubleArrayTopic("orientation").Subscribe({});
}

void Robot::TeleopPeriodic() override {

  // Get the latest value for position and orientation. If the value hasn't been updated since the last time
  // we read the table, use the same value.
  std::vector<double> position = positionSub.Get();
  std::vector<double> orientation = orientationSub.Get();

  std::cout << "position XYZ: ";
  for (double pos : position) {
    std::cout << pos << " ";
  }
  std::cout << std::endl;

  std::cout << "orientation RPY: ";
  for (double angle : orientation) {
    std::cout << angle << " ";
  }
  std::cout << std::endl;
}

DoubleArraySubscriber positionSub;
DoubleArraySubscriber orientationSub;

@Override
public void robotInit() {

  // Add to your robotInit, Command, or Subsystem
  NetworkTable table = NetworkTableInstance.getDefault().getTable("MAPLE");
  positionSub = table.getDoubleArrayTopic("position").subscribe(new double[] {});
  orientationSub = table.getDoubleArrayTopic("orientation").subscribe(new double[] {});
}

@Override
public void teleopPeriodic() {

    // Get the latest value for position and orientation. If the value hasn't been updated since the last time
    // we read the table, use the same value.
    double[] position = positionSub.get();
    double[] orientation = orientationSub.get();

    System.out.print("position XYZ: " );
    for (double pos : position) {
      System.out.print(pos + " ");
    }
    System.out.println();

    System.out.print("orientation RPY: " );
    for (double angle : orientation) {
      System.out.print(angle + " ");
    }
    System.out.println();
}

def robotInit(self):

    # Add to your robotInit, Command, or Subsystem
    table = ntcore.NetworkTableInstance.getDefault().getTable("MAPLE")
    self.positionSub = table.getDoubleArrayTopic("position").subscribe([])
    self.orientationSub = table.getDoubleArrayTopic("orientation").subscribe([])

def teleopPeriodic(self):

    # Get the latest value for position and orientation. If the value hasn't been updated since the last time
    # we read the table, use the same value.
    position = self.positionSub.get()
    orientation = self.orientationSub.get()

    print("position XYZ: ", position)
    print("orientation RPY: ", orientation)

5.4. Updating MAPLE

Navigate to the directory with your docker-compose.yml file in it and run the following command

docker-compose pull