Kinematic Modelling for Mobile Robots

The concept of kinematic modeling is very well studied be it for the AMRs or robotic arms. Being an expert in mobile robotics, I have spent years in the exploration of planners and controllers of differential-drive robots. Although there is always a limitation in the motion range, differential-drive is still the most common kinematic model so far. I have recently started seeing a shift and the adaptation of holonomic robots coming up.
In this blog, I wanted to list the kinematics for both non-holonomic and holonomic robots, along with recommended controllers and planners.

I would use the open-source Linorobot2 package in GitHub, which provides a robust simulation environment complete with Nav2. The sensor configurations I would use for testing the planners are 2D LIDAR, 6-DOF IMU, and Encoders.

Wheel Types in Mobile Robots

Mobile robots come in various configurations, and their movement heavily relies on the type of wheels they use. Let's begin by categorizing the wheels into two main types: Holonomic and Non-Holonomic wheels.

Planners for wheeled robots

In the domain of robotics and autonomous navigation, the choice of planners is a pivotal decision that influences a robot's path-planning capabilities. Whether dealing with holonomic or non-holonomic robots, selecting the right planner is essential. Here, I will delve into the characteristics and applicability of various planners, considering both holonomic and non-holonomic robots.

After conducting thorough tests and measurements with various planners available in Nav2 for non-holonomic robots and holonomic robots, I have gained valuable insights into their performance. Here are my findings based on extensive testing using the Linorobot2 simulation.

Nav2 Path from Start to Goal

For my tests with various planners present in the Navigation Stack in ROS 2, I provided a common navigation goal, which had an approximate distance of 21 meters. I carefully measured the time taken by each planner to complete the given task. Here are the results:

Based on the results, the NavFn Planner appears to be the fastest among the tested planners, taking 92 seconds for non-holonomic robots and 97 seconds for non-holonomic robots respectively. However, its performance may vary depending on the environment's complexity. While all the planners are capable of achieving the navigation goal, the NavFn Planner demonstrated slightly better performance by completing the task in the shortest amount of time (92 seconds).

It's important to note that planner performance can vary based on the specific robot's characteristics, environment, and the nature of the navigation task. Therefore, I recommend further fine-tuning and testing the selected planner to ensure optimal performance in your real-world applications.

Controllers for wheeled robots

Several controllers cater to the diverse needs of both holonomic and non-holonomic robots. These controllers are essential components that help robots navigate through environments efficiently. I will explain the characteristics and suitability of these controllers for different types of robots.

For non-holonomic robots, I would recommend considering the DWB Controller. It's known for its stable performance, simplicity, and relatively fast computation times. It should complement the faster NavFn Planner effectively.

For more advanced trajectory planning and can tolerate slightly longer computation times, the TEB Controller might be suitable, especially if the robot needs to navigate dynamic environments with obstacles.

Given these assessments, the DWB Controller emerges as the top recommendation for holonomic robots in Nav2 due to its remarkable precision and efficiency, which are crucial for responsive and accurate navigation in various environments.


In the ever-evolving realm of robotics, my choice of wheel type, kinematic model, and navigation tools holds the key to unlocking the full potential of my robot. I've embarked on a journey through the diverse landscape of wheel configurations, from the nimble holonomic mecanum wheels to the versatile differential-drive. I've delved into the intricacies of their kinematic models, understanding how they enable robots to move and navigate.

However, my quest doesn't end with theoretical knowledge alone. In the practical realm of robotics, I've put these theories to the test. Through rigorous experimentation, I've identified the NavFn Planner as my swift companion for holonomic robots, while the DWB Controller emerges as the pinnacle of precision for holonomic navigation in ROS 2 Navigation. For non-holonomic robots, the NavFn Planner offers a robust choice, complemented by the agility and obstacle-handling prowess of the TEB Controller.

As I conclude this journey, remember that the world of robotics is dynamic and ever-changing. While these recommendations serve as a compass, my robot's unique characteristics and the complexities of its operational environment will always guide my final decision. So, I'll harness this knowledge, embrace the possibilities, and let my robot embark on its path to innovation and discovery.

If you liked the article, please buy me a ☕️ coffee