SMART AI DRONE

Project Title

SMART AI DRONE

Project Area of Specialization Artificial IntelligenceProject Summary

Introduction

"A drone is a flying robot that may be remotely controlled or fly autonomously using software-controlled flight plans in its embedded systems, functioning in concert with onboard sensors and GPS." Drones may have originated in 1849, when Austria assaulted Venice using unmanned balloons laden with explosives. Around 200 of these incendiary balloons were fired above Venice by Austrian soldiers besieging the city at the time.

Drones, which were initially designed for military use, have witnessed significant development and advances and have made their way into consumer gadgets. Originally, they were used as weapons, in the form of remotely-guided aerial missile launchers. Drones, on the other hand, have a wide range of civilian uses nowadays, particularly in the form of tiny quadcopters and octocopters.

Day by day drones are going to be autonomous for a variety of purposes. Such like drones are used for delivery, surveillance, agriculture etc. An autonomous drone is helpful cause that brings easiness in human’s life to make them each task easy, as it would not need any operator which even wastes his time.

 The milestone of this project is to develope a drone for surveilence purpose. The drone could be used by the armed forces or any government agencies, sepcially working on borders, to detect objects and to track the moveable objects and can save it self from upcoming objects. 

Problem Statement

Always drones are totally dependent on operator: whether that is an agriculture drone, photography drone or any surveillance drone.

Proposed Solution 

To develop a self-functional drone. Which would be able to detect objects, and will track the object automatically when specified. Also, it will save itself from obstacle avoidance.

The drone provides an amazing self-control experience. Person will be able to keep following an object. The object will be pointed out and it would be traced hidden, while the obstacle would be changed on either direction where there is a clean path, the drone will move on it. The project can be more helpful to keep happy the Govt agencies and keep their tasks easy.

Project Objectives

Objectives

This project is to develop a fully self-functional drone, which performs all the functionalities autonomously.

The Functionalities are:

• Object Detection
• Specify an object and Trace it.
• Obstacle Avoidance

i. Object Detection:

As soon as the drone gets started through system. The drone will automatically detect all objects, specially all those objects which are  moveable.  

ii. Object Tracking:

As soon as all the objects detected successfully, the operator can only specify the object from the detected objects. Whichever object has been specified the drone will keep tracking of that object. 

iii. Obstacle Avoidance:

Now, that the drone is following the object which is specified, the drone must save itself from avoiding obstacles. Therefore autonomous self object avoidance functionality has been implemented. 

Project Implementation Method

Methodology

We have chosen prototyping because our drone functionalities will evolve over time and since evolutionary prototyping helps in testing the product in every evolution of the build by showing the result and improving it in next phases.

Since Evolutionary prototyping is to create a model which is refined over time. Like stages of evolution, each stage of the process involves resolving any issues and implementing any changes to the prototype.

In evolutionary prototyping the first build serves as a heart or base for the product which will evolve over time by making changes and doing improvements into the base software.

Software Tools 

i. Pycharm(IDE)

ii. Python (Language)

iii. VS-CODE (IDE)

Hardware Tools :

DJI-TELLO Drone (open-source)

ALGORITHMS

I. YOLO-V4

Yolo V4 is an algorithm which is used for object detection. It will work in two applications here: i. Object Detection and ii. Obstacle Avoidance. In Obstacle Avoidance because it will detect the nearest object also and will change the path.

II. Inception V3

This algorithm works here for tracking or following an object. On the ImageNet dataset, Inception v3 is a commonly used image recognition model that has been proven to achieve better than 78.1 percent accuracy. The model represents the result of numerous concepts explored over time by a number of scholars.

iii. Single Shot Detector ( SSD)

A single-shot detector (SSD) is a detector that detects just one signal at a time. It doesn't use a delegated region proposal network and predicts border boxes and classes from feature maps in a single run. SSD includes tiny convolutional filters to predict object classes and offsets to default border boxes to increase accuracy.

Benefits of the Project

Benefits Of The Project:

I. Security Servilence 

ii. To reduce human dependencies. 

iii. Beneficial for Government agencies. 

iv. To gathered multiple functions to one drone. 

v. To follow or trace a suspecious person. 

Technical Details of Final Deliverable

FINAL DELIEVERABLES:

We have choosen an Open-Source drone which is DJI-TELLO. The final acievement of this project is to develop a self-functional drone. Which would be able to detect objects, and will track the object automatically when specified. Also, it will save itself from obstacle avoidance.

The drone provides an amazing self-control experience. Person will be able to keep following an object. The object will be pointed out and it would be traced hidden, while the obstacle would be changed on either direction where there is a clean path, the drone will move on it. The project can be more helpful to keep happy the Govt agencies and keep their tasks easy.

Final Deliverable of the Project HW/SW integrated systemCore Industry ITOther Industries Security Core Technology Artificial Intelligence(AI)Other Technologies RoboticsSustainable Development Goals Decent Work and Economic Growth, Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	61498
DJI Tello Drone	Equipment	1	50999	50999
Battery	Equipment	1	5999	5999
propellers	Equipment	1	700	700
Proposals Printing	Miscellaneous	4	100	400
Thesis Printing	Miscellaneous	2	1700	3400