Autonomous Multitasking Agent include path planning obstacle avoidance and image processing

Project Title

Autonomous Multitasking Agent include path planning obstacle avoidance and image processing

Project Area of Specialization RoboticsProject Summary

Autonomous Multitasking Agent is a robot that takes decision on its own. It uses stereo camera as its main sensor to generates disparity map and detect  the distance of the obstacle and pass its data to SLAM(Simultaneous Localization And Mapping) algorithm to generate the map of unknown environment. The algorithm takes decisions and deliver the pulses (current) to stm32 Microcontroller to run the omni-wheels.

Project Objectives

The main objective of this project is to implement autonomous mobile agent. The system will use either stereo camera as its main sensor for distance evaluation, path planning, and obstacle avoidance.

Project Implementation Method

Design and implement the robot circuity

Components Required:

• Raspberry pi 4 model B
• 2 Raspberry pi Cameras
• 1 STM32f103c8 microcontroller
• 3 NH5019a Motor Driver
• 3 Quadrature Encoder Motors
• Lipo 3S battery
• USB to Serial TTL module

Raspberry Pi 4 is the main processor to compute SLAM and obstacle avoidance Algorithm. Connection are as follows:

•  Stereo pi (Stereo camera) is connected to raspberry pi to detect obstacle and generate disparity map.
•  STM32f103c8 is connected to raspberry pi 4 via serials.
• 3 Motor Drivers are connected to STM32f103c8.
• 3 Motors are connected to Motor Drivers.
• Motor Encoders are connected to STM32F103c8.

Implementation of SLAM algorithm

The SLAM algorithm is assumed to simultaneously create a map of the mobile robot environment as well as calculating the position within this map. The environmental learning determined by SLAM algorithm of the mobile robotic autonomous tasks. SLAM only has been actively discussed by using cameras because the sensor configuration is simple and the technical difficulties are higher than the others.

Implementation of obstacle avoidance algorithm

Obstacle avoidance method is designed and concern to an experimental autonomous ground vehicle system. This algorithm is easy to tune and it takes into consideration the field of view as well as it’s easy to test in both simulated and real-time of the mobile. Also, it’s very inefficient and therefore various improvements have been proposed.

Benefits of the Project

• The main benefits of this project is its self-driving techniques in transportation.
• Used in restaurants as robot waiters.
• Run in mobile environment without colliding to any obstacle.
• Used as surveillance robot.

Technical Details of Final Deliverable

• Design Omni-wheel Embaded System (Robot circuity) 
• SLAM Algorithms implementation in python.
• Obstacle Avoidance Algorithm implementation in python.
• Autonomous agent run in the moblile environment without collision.

Final Deliverable of the Project HW/SW integrated systemCore Industry TransportationOther Industries Security Core Technology RoboticsOther Technologies Artificial Intelligence(AI)Sustainable Development Goals Industry, Innovation and Infrastructure, Responsible Consumption and Production, Partnerships to achieve the GoalRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	64870
Raspberry pi 4 Model B	Equipment	1	15000	15000
Stm32f103c8 (Microcontroller)	Equipment	1	400	400
Vnh5019 (Motor Driver)	Equipment	3	3000	9000
Quadrature Encoder Motors	Equipment	3	3600	10800
Raspberry Pi V2 camera module	Equipment	2	4000	8000
Stereo Camera HAT for raspberry Pi	Equipment	1	9600	9600
Lipo 3S 12V 28000mah Battery	Equipment	1	2000	2000
IMAX B6 AC 80W(Charger)	Equipment	1	4000	4000
USB to serial TTL module	Equipment	1	270	270
Acralic Sheet	Miscellaneous	1	1000	1000
Holding Rod	Miscellaneous	1	300	300
Omni-Wheels	Miscellaneous	3	1000	3000
Acralic sheets Lazer Cutting	Miscellaneous	1	500	500
Wires (cables)	Equipment	1	1000	1000