Home automation for disabled person using gesture control and artificial intelligence

The most scientifically exploited non-stationary signal acquired from the human brain is electroencephalography (EEG), which plays a key role in brain-computer interface (BCI) applications. EEG signals are linked to brain activities, and BCI technology uses them to create a link between a person's m

Project Title

Home automation for disabled person using gesture control and artificial intelligence

Project Area of Specialization

Artificial Intelligence

Project Summary

The most scientifically exploited non-stationary signal acquired from the human brain is electroencephalography (EEG), which plays a key role in brain-computer interface (BCI) applications. EEG signals are linked to brain activities, and BCI technology uses them to create a link between a person's mental state and a computer-based signal processing system that interprets them. Due to its aperiodic, non-linear, and dynamic nature, decoding an EEG signal with significant variability and non-stationary noise into a meaningful signal is extremely difficult. Non-stationary EEG signals have many frequencies and their amplitudes shift as they propagate through time. The study of such signals necessitates a unique methodological approach and mathematical apparatus that allows for the discovery of the signal's major properties via signal transformation. In this project we have developed a brain gestures (signals) recognition system and machine learning algorithm that controls home automation appliances for physically impaired people. EEG signal analysis for BCI applications is the topic of this dissertation. The BCI's goal is to detect and quantify aspects of EEG signals that represent the user's intents, and then translate these qualities into device commands that fulfil the user's intent in real time. For real-time applications, most classic EEG signal classification algorithms are too complicated. Their performances are insufficient and inefficient in terms of time. In balancing efficiency and accuracy in recognizing tasks from EEG, the existing approaches are also constrained. As a result, building an effective EEG signal decoding system is critical for an improved BCI system to correctly and automatically detect tasks associated to a performed activity which is the key contribution of this project.

Project Objectives

• Develop an embedded system that can aid physically challenged people (deaf, dumb, blind, lame) to control physical devices through brain signals i.e. turn ON/OFF fan or light bulb, open or close the gate, turn ON/OFF television, activate or deactivate an alarm to call someone.

 • Enhance the robustness of the system and reduce wireless communication latency to effectively communicate with applications with any error or delay.

 • This project aims to bring BMI out from laboratories and hospitals to a common man threshold. By doing the comparative analysis with commercially available BMIs, our goal is to make it economical and cost effective.

 • Reduce the size and volume of the device such that it should be convenient for a patient carry it without any extra burden.

 • The device should perform equally well for all types of patients and it must be user friendly. 

Project Implementation Method

. Our work division to implement this project is following:

• The first step is to select a good commercially available EEG sensor.

 • The second step is to train a deep learning model on the data received from the sensor for different types of application.

 • After successfully training the model, interface the sensor and applications with a central processing unit.

• Import the trained model to the central processing unit.

• Train a human brain to efficiently use sensor to control applications.

Benefits of the Project

• Equally effective for all types of patients (deaf, dumb, blind or disable).

• Robust in nature.

 • Cost effective 

• Good quality

 • Better performance

• Should be easy to use and understand

• Plug and play • Not harmful • Electroencephalography based

Technical Details of Final Deliverable

This is a first of its kind project in which a commercially available BMI is used to operate home automation devices for patient’s utility. This robustness of this device is comparable to all other available automation devices. By the use of deep learning algorithms, its robustness and accuracy to predict results has increased. Robust results, decreased size of the device and cost effectivity makes this project ideal for in-home usage for a layman. Moreover, this project is a milestone for future investigations that will be made in the BMI realm. For further research and increased robustness studies, this project will provide a basic framework. As the world is moving towards decreased hardware structures and easy to use interfaces, the potential research opportunity should be operated automation applications using hand gestures that will make this project a zero-hardware based project.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Manufacturing

Other Industries

Education , Health

Core Technology

Artificial Intelligence(AI)

Other Technologies

Others

Sustainable Development Goals

Good Health and Well-Being for People, Industry, Innovation and Infrastructure

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com