GPS Denied Navigation for UAV (Unmanned aerial vehicle )

Project Title

GPS Denied Navigation for UAV (Unmanned aerial vehicle )

Project Area of Specialization Artificial IntelligenceProject Summary

UAV (unmanned aerial vehicle) are usually working along with GPS which is vulnerable to many factors which will be discussed afterwards. The prior or main motive of this project is to shift towards the real time data processing with the help of deep learning and some other sensors like camera for vision based navigation and minimize the use of GPS in our future drones and shift towards more reliable solution. It can be obtained by various methods such as

• SLAM (simultaneous localization and mapping).
• Using camera as a sensor.

Project Objectives

 Drones typically orient themselves using GPS, which helps keep them stable while in flight.

But if you’re flying in a mine—or in a variety of other scenarios such as

1. Power Plants
2. Under the metal bridge
3. Forest.
4. Mining Tunnels.
5. Spoofing & Jamming of signals

               GPS signal may simply not be available.

Solution:  We are going to address this problem in our project  by implementing real time data processing with the help of sensors mainly camera and interfacing it with  the help of deep learning. To depend not on the GPS based

Project Implementation Method

The goal of our system is to enable a UAV/QUADCOPTER to navigate autonomously in corridor environments the capabilities of a human pilot to take reasonable real-time decisions. The images are generated by a front-facing camera attached to the UAV. The trained classifier will return a real-time flight command for safe navigation.

Our training process will enable the UAV to learn most effective strategy to control the navigation with minimum probability of failure. We will use a state-of-the-art CNN model, to train our model. The model parameters are learned by fine-tuning with our custom-made data or predefined paths.

The input to the trained model is a real-time image generated from a camera attached to the front of the UAV. The output of the model will be the different maneuvers done, such as move forward, shift right, shift left and stop.

Main hardware that would possibly be used in the project are jetson nano ,camera for visual navigation ,some flight controllers .

Benefits of the Project

There many benefits of the project in consumer market as well as in coorporate market.

As we know spoofing and jamming of signals is pretty easy these days which can effect the overall performance of the drone eventually.

This project would basically work on the priciple of deep learning.It would be of great use  when flying inside of assets like huge oil storage tanks or industrial boilers, accessing GPS may be really hard, if not impossible.

Mining:Mines present the same kinds of challenges for GPS as flying inside huge above-ground assets—the signal just doesn’t reach.This project can be used in inspection of tunnels and save a lot of human work to manually monitor the overall tunnel.

3.Bridge Inspections: Flying near or under a metal bridge can interfere with your drone’s ability to connect to GPS. Magnetic interference can be caused by steel and electronic objects. Primarily these interferences are caused by power lines, RCC structures, concrete slabs, metal roofs, and steel structures.It is very beneficial  in such conditions.

4.Critical Infrastructure. Some governmental agencies may prefer to use a drone that doesn’t rely on GPS near critical infrastructure, such as military bases or power plants. When operating in sensitive areas, like nuclear power plants,where the visual data collected by drone could directly impact  national security, any kind of connection, be it GPS or internet,  may be seen as a liability.So this project can be widely be used for civil as well as military purposes .

Spoofing And Jamming: As UAVs rely on Global Positioning System (GPS) for positioning and navigation, they can fall prey to GPS jamming and spoofing attacks. The vulnerability of GPS to spoofing has serious implications for UAVs, as victim drones using civil GPS can be misdirected or even completely hijacked for malicious intents.This project eliminates the use of gps making use of drone more safe and reliable.

Technical Details of Final Deliverable

Final deliverable would be a drone;hardware and software integrated.We would be working on constraint requirements of the overall project first and would work on hardware and software integration .e.g jetson nano with camera,jetson nano with controller e.t.c.We would also be working in designing and manufacturing of UAV too.

Experimental evaluation and results in software would also be produced.At last Software UAV integration/flight controller and camera interface would also be done.

Final Deliverable of the Project HW/SW integrated systemCore Industry OthersOther Industries Manufacturing Core Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Industry, Innovation and InfrastructureRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	68000
Jetson Nano 4gb	Equipment	1	32000	32000
Pixhawk 4	Equipment	1	30000	30000
Camera	Equipment	1	6000	6000