Controlling the copter from Arduino

For interaction with ROS topics and services on a Raspberry Pi, you can use the rosserial_arduino library. This library is pre-installed on a Raspberry Pi image.

The main tutorial for rosserial: http://wiki.ros.org/rosserial_arduino/Tutorials

Arudino is to be installed on Clever and connected via a USB port.

Configuring Arduino IDE

To work with ROS and Arduino, you should understand the format of installed packages' messages. For this purpose on Raspberry Pi, build the ROS messages library:

rosrun rosserial_arduino make_libraries.py.

The obtained folder ros_lib is to be copied to <sketches folder>/libraries on a computer with Arudino IDE.

Configuring Raspberry Pi

To run the program on Arduino once, you can use command:

roslaunch clever arduino.launch

To start the link with Arduino at the startup automatically, set argument arudino in the Clever launch file (~/catkin_ws/src/clever/clever/launch/clever.launch):

<arg name="arduino" default="true"/>

After the launch file is edited, restart package clever:

sudo systemctl restart clever

Delays

When rosserial_arduino is used, the Arduino microcontroller should not be blocked for more than a few seconds (for example, using the delay function); otherwise communication between Raspberry Pi and Arduino will be broken.

During implementation of long while cycles, ensure periodic calling the hn.spinOnce function:

while(/* condition */) {
  // ... Perform required actions
  nh.spinOnce();
}

To organize long delays, use the delays in a loop with periodic calling of the hn.spinOnce() function:

// 8 second delay
for(int i=0; i<8; i++) {
  delay(1000);
  nh.spinOnce();
}

Working with Clever

The set of services and topics is similar to the regular set in simple_offboard and mavros.

An example of a program that controls the copter by position using the navigate and set_mode services:

// Connecting libraries for working with rosseral
#include <ros.h>

// Connecting Clever and MAVROS package message header files
#include <clever/Navigate.h>
#include <mavros_msgs/SetMode.h>

using namespace clever;
using namespace mavros_msgs;

ros::NodeHandle nh;

// Declaring services
ros::ServiceClient<Navigate::Request, Navigate::Response> navigate("/navigate");
ros::ServiceClient<SetMode::Request, SetMode::Response> setMode("/mavros/set_mode");

void setup()
{
  // Initializing rosserial
  nh.initNode();

  // Initializing services
  nh.serviceClient(navigate);
  nh.serviceClient(setMode);

  // Waiting for connection to Raspberry Pi
  while(!nh.connected()) nh.spinOnce();
  nh.loginfo("Startup complete");

  // Custom settings
  // <...>

  // Test program
  Navigate::Request nav_req;
  Navigate::Response nav_res;
  SetMode::Request sm_req;
  SetMode::Response sm_res;

  // Ascending to 2 meters:
  nh.loginfo("Take off");
  nav_req.auto_arm = false;
  nav_req.x = 0;
  nav_req.y = 0;
  nav_req.z = 2;
  nav_req.frame_id = "body";
  nav_req.speed = 0.5;
  navigate.call(nav_req, nav_res);

  // Waiting for 5 seconds
  for(int i=0; i<5; i++) {
      delay(1000);
      nh.spinOnce();
  }

  nav_req.auto_arm = false;

  // Flying forward 3 meters:
  nh.loginfo("Fly forward");
  nav_req.auto_arm = true;
  nav_req.x = 3;
  nav_req.y = 0;
  nav_req.z = 0;
  nav_req.frame_id = "body";
  nav_req.speed = 0.8;
  navigate.call(nav_req, nav_res);

  // Waiting for 5 seconds
  for(int i=0; i<5; i++) {
    delay(1000);
    nh.spinOnce();
  }

  // Flying to point 1:0:2 in the marker field
  nh.loginfo("Fly on point");
  nav_req.auto_arm = false;
  nav_req.x = 1;
  nav_req.y = 0;
  nav_req.z = 2;
  nav_req.frame_id = "aruco_map";
  nav_req.speed = 0.8;
  navigate.call(nav_req, nav_res);

  // Waiting for 5 seconds
  for(int i=0; i<5; i++) {
    delay(1000);
    nh.spinOnce();
  }

  // Landing
  nh.loginfo("Land");
  sm_req.custom_mode = "AUTO.LAND";
  setMode.call(sm_req, sm_res);
}

void loop()
{
}

Getting telemetry

With Arduino, you can use the get_telemetry service. To do so, declare it similar to the navigate and set_mode services:

#include <ros.h>

// ...

#include <clever/GetTelemetry.h>

// ...

ros::ServiceClient<GetTelemetry::Request, GetTelemetry::Response> getTelemetry("/get_telemetry");

// ...

nh.serviceClient(getTelemetry);

// ...

GetTelemetry::Request gt_req;
GetTelemetry::Response gt_res;


// ...

gt_req.frame_id = "aruco_map"; // frame id for x, y, z
getTelemetry.call(gt_req, gt_res);

// gt_res.x is copter position on the x axis
// gt_res.y is copter position on the y axis
// gt_res.z is copter position on the z axis

Problem

When using Arudino Nano, RAM may be insufficient. In this case, messages will appear in the Aruino IDE like:

Global variables use 1837 bytes (89%) of the dynamic memory, leaving 211 bytes for local variables. The maximum is 2048 bytes.
Not enough memory, the program may be unstable.

You can reduce RAM usage by reducing the size of the buffers allocated for sending and receiving messages. To do this, place the following line at the beginning the program:

#define __AVR_ATmega168__ 1

You can reduce the amount of used memory even more, if you manually configure the number publishers and subscribers, as well as the size of memory buffers allocated for messages, for example:

#include <ros.h>

// ...

typedef ros::NodeHandle_<ArduinoHardware, 3, 3, 100, 100> NodeHandle;

// ...
NodeHandle nh;

results matching ""

    No results matching ""