Elderly Monitoring Application Through Human Activity Detection

Project Title

Elderly Monitoring Application Through Human Activity Detection

Project Area of Specialization Artificial IntelligenceProject Summary

Statistics show that the population of aging people is increasing and in the near future, human assistance for elderly people will be expensive because the ratio of caretakers is less than the ratio of the elderly population. Hence, we have proposed a system that will be able to monitor the activity of the elderly through mobile motion sensors. This project is concerned with identifying the specific movement or action of a person based on sensor data by its classification into one of the training classes. The system will be able to  intelligently detect when an older adult is passive or moving e.g., take a walk, sit, fall, etc.

The application will provide two modules, one for the patient and the other for the caregiver. When anomalous activity is detected the phone will notify the caregiver with a notification over his device. Multiple sensor available in mobile phones such as accelerometer, gyroscope, gravity sensor etc can be used for the proposed system. MobiFall dataset will be used in the prototype for testing of the system. 

Project Objectives

The main objective of that the project will achieve are:

• Preprocessing of collected samples using data mining techniques
• Selecting a classification algorithm for application after testing data on test samples
• Provide an android application to monitor elderly patients movement in real-time
• Provide notification on care-giver mobile in case of fall detection

Project Implementation Method

The project shall involve the following steps for implementation: 

1. Data collection
2. Load the dataset
3. Preprocess the data
4. Training the dataset
5. Feature selection and extraction
6. Learning and classification
7. Recognize activity and evaluate the result
8. Integration in the android application

First of all, we shall work on the logic layer and collect the data input from the mobile phone. Than we load the dataset into a python based model. Dataset is the collection of data pieces that can be treated by computer as single unit for analytic and prediction purpose. To do preliminary processing of collected data we preprocess it. After that we train the dataset through machine learning algorithms. For supervise machine learning we use labelled dataset, so that machine can easily and clearly understand the input patterns than we train the modules and test them. After all the process we made the final model so through that we demonstrate the results.

Benefits of the Project

Our proposed system detects dangerous activities such as falls to provide necessary help in time.

Technical Details of Final Deliverable

The final deliverable will be an androud application with an integrated AI module. 

Android application:

• • Android APIs: Various android APIs will be used in the project for varying purposes such as recording sensor input, user interface, tensor flow model integration etc.  SDK: provides the API libraries to use the sensors.
  • IDE: An integrated development environment (IDE) is software for building applications that combines common developer tools into a single graphical user interface (GUI)
  • Language:  Front-end: XML, Back-end: java

Artificial Intelligence Module:

• • Python Libraries and Frameworks: Python implementation of KNN and LSTM  classification algorithm.
  • Google Colab: An online IDE for implementation of python code using google drive

Final Deliverable of the Project Software SystemCore Industry ITOther IndustriesCore Technology Artificial Intelligence(AI)Other TechnologiesSustainable Development Goals Good Health and Well-Being for PeopleRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	65000
GeForce GTX 1070	Equipment	1	65000	65000