Exploring ROS2 with a wheeled robot – #4 – Obstacle avoidance

06 December 2021

share this:

By Marco Arruda

In this post you’ll learn how to program a robot to avoid obstacles using ROS2 and C++. Up to the end of the post, the Dolly robot moves autonomously in a scene with many obstacles, simulated using Gazebo 11.

You’ll learn:

  • How to publish AND subscribe topics in the same ROS2 Node
  • How to avoid obstacles
  • How to implement your own algorithm in ROS2 and C++

1 – Setup environment – Launch simulation

Before anything else, make sure you have the rosject from the previous post, you can copy it from here.

Launch the simulation in one webshell and in a different tab, checkout the topics we have available. You must get something similar to the image below:

2 – Create the node

In order to have our obstacle avoidance algorithm, let’s create a new executable in the file ~/ros2_ws/src/my_package/obstacle_avoidance.cpp:

#include "geometry_msgs/msg/twist.hpp"    // Twist
#include "rclcpp/rclcpp.hpp"              // ROS Core Libraries
#include "sensor_msgs/msg/laser_scan.hpp" // Laser Scan

using std::placeholders::_1;

class ObstacleAvoidance : public rclcpp::Node {
  ObstacleAvoidance() : Node("ObstacleAvoidance") {

    auto default_qos = rclcpp::QoS(rclcpp::SystemDefaultsQoS());
    subscription_ = this->create_subscription(
        "laser_scan", default_qos,
        std::bind(&ObstacleAvoidance::topic_callback, this, _1));
    publisher_ =
        this->create_publisher("cmd_vel", 10);

  void topic_callback(const sensor_msgs::msg::LaserScan::SharedPtr _msg) {
    // 200 readings, from right to left, from -57 to 57 degress
    // calculate new velocity cmd
    float min = 10;
    for (int i = 0; i < 200; i++) { float current = _msg->ranges[i];
      if (current < min) { min = current; } } 
    auto message = this->calculateVelMsg(min);
  geometry_msgs::msg::Twist calculateVelMsg(float distance) {
    auto msg = geometry_msgs::msg::Twist();
    // logic
    RCLCPP_INFO(this->get_logger(), "Distance is: '%f'", distance);
    if (distance < 1) {
      // turn around
      msg.linear.x = 0;
      msg.angular.z = 0.3;
    } else {
      // go straight ahead
      msg.linear.x = 0.3;
      msg.angular.z = 0;
    return msg;
  rclcpp::Publisher::SharedPtr publisher_;
  rclcpp::Subscription::SharedPtr subscription_;

int main(int argc, char *argv[]) {
  rclcpp::init(argc, argv);
  return 0;

In the main function we have:

  • Initialize node rclcpp::init
  • Keep it running rclcpp::spin

Inside the class constructor:

  • Subcribe to the laser scan messages: subscription_
  • Publish to the robot diff driver: publisher_

The obstacle avoidance intelligence goes inside the method calculateVelMsg. This is where decisions are made based on the laser readings. Notice that is depends purely on the minimum distance read from the message.

If you want to customize it, for example, consider only the readings in front of the robot, or even check if it is better to turn left or right, this is the place you need to work on! Remember to adjust the parameters, because the way it is, only the minimum value comes to this method.

3 – Compile the node

This executable depends on both geometry_msgs and sensor_msgs, that we have added in the two previous posts of this series. Make sure you have them at the beginning of the ~/ros2_ws/src/my_package/CMakeLists.txt file:

# find dependencies
find_package(ament_cmake REQUIRED)
find_package(rclcpp REQUIRED)
find_package(geometry_msgs REQUIRED)
find_package(sensor_msgs REQUIRED)

And finally, add the executable and install it:

# obstacle avoidance
add_executable(obstacle_avoidance src/obstacle_avoidance.cpp)
ament_target_dependencies(obstacle_avoidance rclcpp sensor_msgs geometry_msgs)



Compile the package:
colcon build --symlink-install --packages-select my_package

4 – Run the node

In order to run, use the following command:
ros2 run my_package obstacle_avoidance

It will not work for this robot! Why is that? We are subscribing and publishing to generic topics: cmd_vel and laser_scan.

We need a launch file to remap these topics, let’s create one at ~/ros2_ws/src/my_package/launch/

from launch import LaunchDescription
from launch_ros.actions import Node

def generate_launch_description():

    obstacle_avoidance = Node(
            ('laser_scan', '/dolly/laser_scan'),
            ('cmd_vel', '/dolly/cmd_vel'),

    return LaunchDescription([obstacle_avoidance])

Recompile the package, source the workspace once more and launch it:
colcon build --symlink-install --packages-select my_package
source ~/ros2_ws/install/setup.bash
ros2 launch my_package

Related courses & extra links:

The post Exploring ROS2 with a wheeled robot – #4 – Obstacle Avoidance appeared first on The Construct.

The Construct Blog

Related posts :

How drones for organ transportation are changing the healthcare industry

The healthcare drone industry has witnessed a dramatic surge in the last couple of years. In 2020, the market grew 30% and is expected to grow from $254 million in 2021 to $1,5 billion in 2028.
21 March 2023, by

Robotic bees and roots offer hope of healthier environment and sufficient food

Miniature robots that mimic living organisms are being developed to explore and support real-life ecosystems.
18 March 2023, by

Robot Talk Episode 41 – Alessandra Rossi

In this week's episode of the Robot Talk podcast, host Claire Asher chatted to Alessandra Rossi from the University of Naples all about social robotics, theory of mind, and robots playing football.
17 March 2023, by

Mix-and-match kit could enable astronauts to build a menagerie of lunar exploration bots

Robotic parts could be assembled into nimble spider bots for exploring lava tubes or heavy-duty elephant bots for transporting solar panels.
14 March 2023, by

Learning to compute through art

“Introduction to Physical Computing for Artists” at the MIT Student Art Association teaches students to use circuits, wiring, motors, sensors, and displays by developing their own kinetic artworks.
12 March 2023, by

Robot Talk Episode 40 – Edward Timpson

In this week's episode of the Robot Talk podcast, host Claire Asher chatted to Edward Timpson from QinetiQ all about robots in the military, uncrewed vehicles, and cyber security.
10 March 2023, by

©2021 - ROBOTS Association


©2021 - ROBOTS Association