Michele Galullo
Leonardo
May 21, 2023
Period
- 2018
Title
I used this software to simulate a movement of the segway on the floor. In particular, for each wheel I simulated the rolling motion. Gazebo was very useful to me as it helped me improve the control algorithm before testing it on a real segway.
Pro
It is easy tool to use with ROS. You can create your own simulation environment in order to avoid fails with the real hardware.
Con
The documentation is not very thorough. During my tests it often crashed. The graphical interface is not intuitive to use.
Robot
Frank a prototype of AlterEgo wheeled humanoid robot.
Antonio Manuele
Università di Pisa
May 20, 2023
Period
- 2020
Title
I used GAZEBO for two different applications. The aim of the first application was the simulation of a humanoid robot, that allowed me to verify the correct implementation of the kinematic simulating the physical characteristics of the real robot. The other application was to simulate a manipulator.
Pro
It allows to simulate a lot of physical phenomena such as gravity, friction, lights and shadows. It is also possible to assign certain torques to the joints in order to simulate the maximum torque of the motor used in the real application.
Con
Sometimes it crashes. Create a robot from zero using XML is not easy at all.
Stefano Maugeri
Intecs
May 6, 2023
Period
- 2022
Title
I worked with GAZEBO to simulate a bipedal robot.
The aim was to verify the functioning of the walk.
GAZEBO allowed me to verify how the control algorithm could behave in a real environment, simulating the terrestrial physics.
GAZEBO is an intermediate tool, to be used after the validation of the simulation control algorithm but before testing it on the real robot.
By simulating the real robot in GAZEBO's simulated physics it is possible to get an idea of how the real robot will behave.
Pro
Gazebo is an open-source software, usable for free by companies.
It is also maintained by a huge community, which continuously improves and updates it.
GAZEBO is supported by ROS, which means it can be used in any area of robotics (manipulator, automotive, drones, robots...).
Furthermore, thanks to ROS, GAZEBO can be interfaced with different programming languages, such as C++ and Python.
Internet is full of both GAZEBO documentation and already solved problems by other users.
Con
GAZEBO is mainly supported by linux. The Windows version is not as up-to-date as the Linux version.
The official documentation sometimes contains errors, or even links to non-existent URLs.
GAZEBO, being mainly used via ROS, is therefore usable mainly on linux, and only with specific (old) versions (eg: ubuntu 18.04)
Creating robots using XML files may not be easy for everyone, it's definitely not the easiest method.
Sometimes, unfortunately, the application crashes.
Robot
Company products
Mathieu Lambert
VAN MELLAERT
Feb 3, 2023
Period
- 2019
Title
Pre-sale simulation of lawn mower robots. Studies of the security of the product and the interaction/collision with the various possible obstacles (trees, animals, etc.)
Pro
It helps to create a dynamic simulation using multiple high-performance physics engines. It allowed me to simulate robots with dozens of sensors without problems.
GAZEBO is very easy to use and can be taught to beginners very easily, it is open source which is a really important point.
Con
It's really greedy, and what's unfortunate is that it doesn't use my graphics card but only the CPU, which doesn't make the task any easier. It consumes a lot of resources in terms of memory and CPU compared to V-REP.
Sofia Faraci
Università di Pisa
Jan 31, 2023
Period
- 2022
Title
I used this tool for two different projects. The first time to test a force estimation algorithm for a wheeled humanoid robot and later to test a Teaching by Demonstration learning system applied to a manipulator robot. Gazebo is a tool. It allowed me to make adjusting to my algorithms before testing them on real robots.
Pro
It's an open-source tool. It's one of the easiest tools to use with ROS. It's a good way to test your work when a physical robot is not available. You can create your own testing environment by using the URDF of the objects.
Con
It often freezes or shuts down even when running on a high-performance laptop. The documentation is not exhaustive.
Robot
Franka Emika Panda manipulator robot and AlterEgo wheeled humanoid robot.
Mariangela Menolotto
Intecs Solutions
Jan 12, 2023
Period
- 2021
Title
I used Gazebo to test a control law for allowing a robot to approach variable slopes. The robot was a humanoid robot over a unicycle with two soft arms and hands. The environment and the robot were provided to me by my university, so I only had to apply the control law (programmed in C++). The inputs to the motors (two DC motors that controlled the wheels) were passed to the simulation through ROS. The available sensors were an IMU fixed to the middle point of the wheels axis and one encoder for each wheel.
Pro
It is open-source and the documentation and many projects can be found online. It is also fairly easy to use.
Con
The simulations were often slow and non-repeatable (maybe because of the soft components), which was the reason why I ended up using other tools to simulate the system. It is also a very heavy program.
Robot
Alter-Ego, a humanoid mobile robot developed by Centro Piaggio of Università di Pisa, Italy.
Mohan Raj Nadarajan
SR Privtae Limited
Dec 4, 2022
Period
- 2022
Title
Simulate the environment of the robot to study the trajectory plans and object manipulation of the robot in pick and place application and check the performance of the robot when a dynamic obstacle pops up in the pathway. The performance of local planner and global planner is studied using Gazebo. In addition the study feed forward estimation of planner is studied for dynamic obstacles.
Pro
This helps to work offline, simultaenously during the development of the robot. Simulation of dynamic obstacles is a very good feature to test the behaviour of algorithms. Importing of robot model in URDF plugins, it allows robot configuration to changed and just import. The gazebo is a physic engine to simulate the robot which eventually delivers the task time required and some simulations.
Con
The package is very huge and it sometimes makes the whole of the host system slower. The booting up time for the application is really high. It is not possible to test something quickly, so on the fly testing is very to do. The integration of gazebo is not very user friendly meaning that there are no much APIs available to integrate with modern programming language frameworks.
Robot
KUKA youBot
Sanket Palchaudhary
SVIT Vasad
Nov 10, 2022
Period
- 2021
Title
Simulated a pick and place robot to pick an object from marker 1 and place it on marker 2. Used a TMR 700 robot from Techman. Tested the robot on various different levels and with changed obstacles to test the working and efficacy of the system.
Pro
The product is very easy to use and can be taught to beginners very easily. It is great for implementation of beginner level projects and has a lot of advanced features which can be used to simplify and control simulations. The interface is user friendly and there are multiple documentations available online to help navigate through almost all issues faced. It is open source which makes it highly accessible to everyone irrespective of barriers. It also has great interconnectivity with other apps and therefore making the work of rookie engineers much easier.
Con
The product requires optimum hardware condition and can consume high RAM while running a high-end simulation which may cause heating issues in system.
Robot
Techman TMR700