Dynamic Obstacle Avoidance Using Mobile Robots

Project Title

Dynamic Obstacle Avoidance Using Mobile Robots

Project Area of Specialization RoboticsProject Summary

The project involves the P3-DX mobile robot also known as the Peoplebot on which the project’s data and testing is carried out. The project focuses on developing a good dynamic obstacle avoidance algorithm to be implemented on the Peoplebot.

The project works in two domains namely software and hardware. Software part of the project works on developing the algorithms on the Robot operating system (ROS), which is a reliable operating system to work with robot, and on GAZEBO, which is robot simulator that works in collaboration with ROS to simulate the algorithms on the robot in a simulated environment. After the algorithm is successfully tested in a simulated environment, the next step is to test the algorithm in an actual environment.

The environment is mapped by the robot using its IR, LIDAR sensors and the camera’s input. The robot also uses these sensors to navigate its path while moving in a dynamic environment where there are moving obstacles is its path of motion.

Project Objectives

The project’s main goal is to develop an algorithm that senses and avoids static and moving obstacles while moving in a mapped environment from one defined position to another. In short terms, the project aims to test and run dynamic obstacle avoidance algorithms on the P3-DX Peoplebot. The algorithm aims to use the Peoplebot’s own sensors and camera to map the environment and also navigate through its mapped environment. The development of our project’s indigenous algorithm involves testing and comparing the already existing obstacle avoidance algorithms.

Project Implementation Method

The robot that we will be using for our testing of the dynamic obstacle avoidance algorithms is a P3-DX PeopleBot.

We will be implementing ROS with our Robot P3-DX. ROS framework is run by a Master server which is usually our Laptop. Our robot will be a node in this whole framework which we could then publish and subscribe to in accordance with the code we make. The robot will be connected to a computer through a RS232 USB port. ROS packages will be used to have advanced control over the robot as well. All the sensors including: 2-D LIDAR, RGB-D camera, Sonar Sensors, will be connected to our ROS framework through different packages. Then, they will be converted to nodes that can be published or subscribed to, along with, our robot, which will also be a node in our framework. Messages can then be sent to each of these nodes individually for modularity and simplicity in our obstacle avoidance algorithm.

Along with the hardware, we will be testing our algorithms on the software side first. For safe and practical testing of robotic system in progress, a well-developed simulation environment is to be used. Gazebo works along with ROS.  Accurate models of the simulated robot and its working environment are to be designed on Gazebo for robot navigation. DIfferent ROS nodes are created to simulate the bot on gazebo. A ROS package called GMAPPING is used for Mapping. RViz and RQT softwares are used for GUI of this whole process. A node for Dynamic Obstacle avoidance will be created in the upcoming weeks.

Benefits of the Project

PeopleBot can avoid obstacles while navigating inside the lab, on flat ground or a slight slope. It can be used to assist people in terms of everyday tasks such as mobility aid, pick-up of objects. This mobile research robot is the ultimate human-robot interaction champion. PeopleBot can operate in social environments and domains, such as assisting the elderly and severely disabled people.

This obstacle avoidance algorithm that we will design can be used in Autonomous Vehicle Navigation, Indoor robot movements, Drones. It will help build the infrastructure to a better autonomous navigation system.

Technical Details of Final Deliverable

On the software side our final deliverables would be:

1. Simulation Enviornment on Gazebo
   We will create the simulation enviornment on Gazebo alongwith the URDF of our PeopleBot which can be used by anyone who has to work on this project in future.
2. Algorithm Devolopment for Dynamic Obstacle Avoidance
   Algorithm is to be developed to carry out dynamic obstacle avoidance
3. ROS Nodes for PeopleBot and its sensors will be created which can be used by anyone for future projects
4. Obstacle Feature extraction will be done by using PeopleBot's camera

On Hardware Side, our final deliverables would be:

1. Integration of ROS with Red Hat Linux
   ROS will be installed on PeopleBot and will be connected remotely to our laptop for user friendly enviornment.
2. A camera will be installed on PeopleBot
3. Testing verified simulation results on PeopleBot
   Tested and verified algorithms will be carried out on PeopleBot to avoid damage and increase our work efficiency.
4. Effiecient Path Planning
   After completing our project, PeopleBot will be able to carry out effiecient path planning. This code can be used on any bot with some minor modifications.

Final Deliverable of the Project Hardware SystemCore Industry TransportationOther Industries Manufacturing Core Technology RoboticsOther Technologies Artificial Intelligence(AI)Sustainable Development Goals Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	60997
Rasberry Pi	Equipment	1	12000	12000
LIDAR	Equipment	1	33000	33000
IR Sensors	Equipment	5	1200	6000
Wires and Components	Miscellaneous	1	9997	9997