Scout Rover

Project Title

Scout Rover

Project Area of Specialization RoboticsProject Summary

We are envisioning a system where an Unmanned Ground Vehicle (Rover) is able to autonomously navigate in an unknown indoor environment and generate a 2D map at realtime. We aim to develop this system cost-effectively by using only visual sensors. The rover shall capture visual input and send it back to a server. The server shall process the visual input by implementing SLAM algorithms. Based on this processing, the server will then autonomously control the rover to navigate and explore the unknown environment. Meanwhile, the server will also generate the 2D Map which shall be visible on a screen.

Project Objectives

1. Set up the Rover hardware: 4 DC Motors, Arduino chip, RaspberryPi, and Kinect Vision Sensor (all mounted on the rover).

2. Set up a Server on some Cloud Platform/Local Machine.

3. Establish communication channels between Rover and Server:

   1. Arduino - RaspberryPi Bluetooth (RFCOMM)

   2. RaspberryPi - Server WiFi (TCP/IP)

4. Implement SLAM algorithms and techniques on the server:

   1. Localization

   2. Motion Planning

5. Generate and display 2D Map on some GUI.

Project Implementation Method Hardware Components

Our rover has both the Arduino and Raspberry Pi. Arduino controls motor driver - which is responsible for powering wheel motors. The connection between Raspberry Pi and Arduino is established through Bluetooth using HC-06 module.

Raspberry Pi gets video feed from the Visual Sensor which is also mounted on the rover and has a wired connection with the Raspberry Pi. This feed is forwarded to the Server through Wifi where all the computations are done.

Software Components

As we have a distributed system, different software modules are running on different hardware components. The Server is responsible for running the main SLAM algorithms while the Raspberry Pi only forwards the video feed to the Server.

Referring to the Class Diagram, Rover class is running on the Raspberry Pi and the Server class is running on the Server. NetworkHandler is a common class which is deployed on both systems.

For the implementation of SLAM Algorithm, Map, Landmark, Particle, Localization and are running on the server side. The landmark class holds the information about the predefined information about the landmarks. Map class translates the environment into classes. Particle class holds the probability of location of the rover. Localization class holds the algorithm for localizing rover against the provided environment map.

  Class Diagram

Benefits of the Project

Although we are only envisioning a proof of concept for the scope of this project, there are tons of opportunities in the real world in which such a system could help improve human life. These are not only business opportunities but also provide a chance to make some positive impact for humanity. Some of these opportunities include:

• Search and Rescue Missions

• Assistant for Vision Impaired People

• Automated Map Generation in Game Development

Technical Details of Final Deliverable

Although the system shall be scalable, however, for the purpose of this project, we limit the scope to only the following features:

• Indoor Environments only

• Single Floor i.e., rover shall be able to move on one plane only

The following constraints will not be handled in this project:

• Poor Wi-Fi

• Low Battery

• Bluetooth range

• Bad Lighting Condition

Final Deliverable of the Project HW/SW integrated systemCore Industry ITOther IndustriesCore Technology RoboticsOther TechnologiesSustainable Development Goals Good Health and Well-Being for People, Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	28500
Rover	Equipment	1	6000	6000
Raspberry Pi	Equipment	1	9000	9000
ESP 32	Equipment	1	1200	1200
Microsoft Kinect	Equipment	1	7000	7000
Lipo 12V battery	Equipment	2	2500	5000
Motor Driver	Equipment	1	300	300