COVID-19 Prediction and Crisis Assistant

For better response and planning for COVID-19, we are aiming to develop a unique and multipurpose surveillance system (for both mobile and web) using big data analytics and machine learning technologies. This platform will provide three different modules such as 1) predication and forecasting, 2) task dispatch and assignment application and 3) volunteer activity management. The first product is the data analytics engine that will be developed for prediction and forecasting purposes. This can be expanded out into the future to provide data visualization and predictions on top of curated open and crowdsourced data. Secondly, the application provides benefits by identifying those in need with those who wish to help. Additionally, it benefits the government and private sector agencies to reach citizens and customers. Thirdly, a volunteer-based database is crowdsourced containing information on both paramedics and general people who are willing to help for social working and government health authorities. 

Predication and forecasting methods about COVID-19 diseases caused by the novel Coronavirus globally need to be developed using state-of-the-art and interactive visualization methods so that various prediction and forecasting models can be applied for better understanding. Moreover, for the visualization of data, looking for trends, or statistical evaluation all the data must be fed to any statistical evaluation state-of-the-art platform readily available for general public to interact and provide feedback. If the data will be fed directly to the database by the paramedical and other regulatory authorities and in a matter of seconds the data is accessible to millions and all law enforcement agencies with accurate predictions and forecasting regarding its spread and intensity. Moreover, information on the socio-economic factors of the people who are affected by the pandemic needs to be gathered through crowdsourced data for better understanding and effective logistic distribution purposes. 

State-of-the-art technology needs to be used by both citizens and governmental personnel providing both parties with what they need, such as a volunteer database allowing humanitarian logistics with the efficient allocation of medically trained reserved personnel, or for citizens by connecting high-risk groups with those who are able to provide food or medicine deliveries. 

The main objectives of this project can be divided into the following three components.

1. To develop a platform that uses state-of-the-art big data analytics and machine learning algorithms to predict and estimate the potential new affected areas, identify vulnerable populations and key factors that are contributing to the spread of the disease. Forecast infection rates and spread intensity to vulnerable populations so that hospitals and health authorities can effectively plan resourcing and response. Moreover, the proposed system will provide interactive statistical visualization to show the spread of the diseases overtime at various geographical locations based on crowdsourced data from recognized sources and government press releases.

2. A task distribution application that registers daily wage service providers with task support application for the registration of tasks related to construction, maintenance and repairing (i.e., electrician, plumber, cleaner etc.) and assigns them targets defined through the requests initiated by people in need.

3. An interactive database containing all the required information about the paramedics and volunteers willing to assist the health authorities. This database can be integrated with any health care application to provide data on the locations, tasks and duty reports of the medical staff for effective management of the personnel.

The project envisages widespread use by both citizens and governmental personnel providing both parties with what they need, such as a volunteer database allowing humanitarian logistics with the efficient allocation of medically trained reserved personnel, or for citizens by connecting high-risk groups with those who are able to provide food or medicine deliveries. Additionally, we will provide a platform for construction, maintenance and repairing service providers to be assigned to the nearby homes and offices requesting specific daily waged services. The information such as details of work, location and budget will be provided to service providers (i.e., electrician, plumber, cleaner, food delivery etc.) for further consultation. 

The project is of great importance for all the stakeholder including government institutions, public and private institutions for their effective preparedness in a timely fashion in case of pandemic and other crisis. The timely response to such pandemic, shall minimize its damage in terms of lives but also economic domain. Moreover, the on-demand home service providers with daily wages model and users requesting can be key stakeholders as a percentage of service can be deducted to improve the business model.

The project once completed can provide great assistance to public health and concerned government and non-government authorities. Moreover, citizens and service providers can benefit from information dissemination, request generations and task allocations components of the proposed system. Sustainability of the application is ensured by focusing on ecosystem development, not just the technology. In the long term, the solution has potential use cases in governmental disaster response, aid distribution, on-demand service task management, volunteer allocation, and epidemiological predictions. 

The project is based on Big data analytics and machine learning technologies. Big Data Analytics is known to facilitate both batch and real-time processing of data streams and is capable of providing meaningful results for understanding the situations persisting in the disaster-affected areas, hence based on the analytical results the deployment of resources is optimal and effective. Moreover, big data analytics provide an opportunity to analyze datasets generated for almost any source, hence providing a larger scope for performing data analytics, monitoring, forecasts and generating alerts for unusual events. Therefore, we argue that the joint exploitation of Big Data Analytics techniques and Machine Learning algorithms can lead to the development of an innovative, effective and highly needed crisis prediction and forecasting environment.

The platform will gather data in a centralized way (via multiple channels, such as: official statistics, crowd sourced data, social media, or self-reported information) allowing for easy generation of statistical reports and data visualization, it is planned to make this data available via application programming interface (API) to ensure widespread use, interoperability, and transparency.

The potentials in our proposed idea will serve as a catalyst to revolutionize the human resource management for the pandemic and other healthcare-related crisis. We strongly believe that the proposed application will make accurate predictions and the application will be a state-of-the-art tool to manage healthcare-related tasks and provide an interactive platform to match the requirements of for people in need

Hadoop ecosystem along with Spark engine is proposed for big data analytics. Decision trees supported with GraphX technology of the Hadoop ecosystem are going to be deployed for the proposed system. Predictive models using Neural networks and Kalman filters are under consideration depending on the datasets we acquire. For classifiers in machine learning, we are planning to work with time Series and clustering algorithms. We are already working on similar technologies and projects so with current expertise level it is easy for the team to execute research in respective domains mentioned in the proposal. Furthermore, the big data analytics tools are open source which means they do not require any licensing fees.

1. https://www.youtube.com/watch?v=7t75CNC34vU

2. https://www.youtube.com/watch?v=HrKzXLgGohA

As COVID-19 continues to spread, the relevant health authorities, governmental organizations, private sector organizations, volunteers and citizens are continuing to face rapidly evolving challenges, such as: lack of updated information, missing data of the required personnel with the required skill sets, inadequate communication capabilities, and inaccurate or incomplete field data. The combination of these leads to a situation where the authorities and the general public are trying to fight and contain the virus from a deprived position.

There is a big scope in providing forecasts and predictions related to the spread of diseases on defined geographical locations backed by state-of-the-art big data analytics and advanced machine learning giving interactive and easy to understand visualization. The proposed project as an additional module aims to enable these people to create a list of groceries and pharmacy items in advance and post a request for nearby volunteers to help. The system will provide a clear overview of the crisis and the information of where assistance is required. The overall task management that encapsulates problems like finding the nearest available option and searching for the desired details does involve ML and BDA-based research in most of the cases, hence making a connection with the setup of the first module. Additionally, the system will register the volunteers and will let them know when and where their skills are needed. Within the context of volunteering works, there are thousands of qualified medical personnel who are willing to contribute, but it is impossible for many health care organizations to keep track of them and recruit them for the required operations. 

The focus of this project is to address the challenges faced during gathering data from multiple channels, integrating it into a single application, and using the gathered data to provide citizens and government officials with accurate statistical information, whilst, at the same time, providing new functionalities such as epidemiological forecasting and prediction, outbreak location detection, disease geo spread, gathering information on medical personnel and volunteers and assigning tasks to them for medical assistance and providing a platform for on-demand home services for quarantine requirement adherence respectively.

Balochistan University of Information Technology, Engineering and Management Sciences has given a new vision and a distinct meaning to education. Its guidelines for academic achievements are comparable to those of the most renowned institutions of the world. At BUITEMS, we are a community of professionals committed to grooming our students for their professional lives.

18

We are experienced and have the required skills. 

This proposal aims to collaborate with stakeholders from both the public sector and citizens at large throughout the design and implementation process, in a co-creative manner. Thus, the overarching methodology for this proposal is a combination of systems thinking and action research. Systems thinking provides a framework for thinking about hard problems and framing them in a way that allows the researcher to understand the system’s emergent behavior. In tandem with this, action research and agile methods have both proven themselves when it comes to solving specific and hard challenges through an iterative process that places a high level of importance on involving outside stakeholders into the innovation process. Thus, the proposed project will start by launching an initial MVP (minimal viable product) of the service within the first month of the project as an MVP due to the nature of the requirements keeping in view the urgency caused by the spread of COVID-19. The development cycle will be iterative, with each development cycle lasting 2 months and a final development cycle of one month. Thus, releases would be expected at the end of Month (M) 1, end of M3, end of M5, and end of M6 for a total of 4 releases. Throughout the process, all relevant stakeholders will be involved. This involvement occurs through multiple channels, such as testing the tool and providing feedback, directly contributing to our open-source code base, collecting and submitting data, or in a passive manner through social media usage.

For the business plan, once we have gained a large user base, we can aim to bring on more private sector companies and individual service providers to our application to shift to services like electrician, plumber, food delivery, etc., Secondly, we can market the data visualization and machine learning components that work behind the scenes. For revenue generation we are targeting the following four different streams of income for our app that can span more with further usage;

• Google AdSense: Google AdSense is a program run by Google that allows publishers in the Google Network of content sites to serve automatic text, image, video, or interactive media advertisements, that are targeted to application content and audience. We are aiming to monetize the free version of our application and its contents using the Google AdSense program.
• App purchasing from Google PlayStore: Since the free version of our application will contain Google Ads., we will offer a paid version of our application to users that will be ads free.
• Registration fee from enlisted businesses: Once the development of our proposed application is completed and a fully functional version is launched, we will then charge a certain amount of money from all the new businesses listed on our platform. Note that every business listed on our platform will first be scrutinized and verified by our marketing team.Paid promotions of enlisted businesses: Revenue will also be generated from running paid contents and promotions (e.g. via push notifications) in our application.  

For research funding.