Let’s move by “deconfinig”!

SUMMARY

We are so worried for the actual emergency situation, we have realised it is very difficult for our goverments to make decisions without the proper tools, which could assure them they are acting correctly.

Our tool aims at providing guidelines for decision-making and management of the de-escalated confinement process. The objective is to help governments/administration in the decision making to elaborate a consistent strategy to ensure getting out correctly from confinement. The “de-confinement” process should be handled carefully in order to avoid new outbreaks and to manage the medical and other critical resources.

The solution we bring to the table:

By means of a number of selected KPI (key performance indicators) based on data analytics and machine learning, we will provide real time status to help in the decision-making process during the different phases of the de-escalated process of the confinement.

The KPI (key performance indicators) have been intentionally and conscientiously selected. By closely monitoring those indicators it is possible to have a correct follow up to contain and gradually solve the crisis.

The parameters (data) that feed the KPIs have been classified in clusters that can be easily modified, added or removed depending on the evolution or their suitability.

Additionally, the set-up allows to extract the information/indicators by regions in order to help in the de-escalated confinement according to the status of the different places.

Our proposal is to control the de-confinement in a bottom up way, allowing gradual people’s movement in a free way but obviously by respecting social distance and the known “good-practices”.

The application will warn the authorities, depending on the KPI evolution, anticipating if it is necessary to revert the measures and the people’s movement.

The CITIC of University of A Coruña is a public institution that can provide the Hardware to host safely both the data and the application software

This project has a strong added value to solve the crisis and recover social and economic activities in a controlled way by reducing and controlling the risk of contagion and the capability of the health&sanitary system to absorb the potential infected people until a vaccine will be available.

Our intention is to develop a powerful tool to learn from the current pandemic, helping us to understand better its evolution and to improve the management of future health threats.

                                   THE PROJECT DEVELOPEMENT:

The problem your project solves

• This project aims at providing guidelines for decision-making and management of the de-escalated confinement process.

• The objective is to help governments/administration to elaborate a consistent strategy to ensure getting out correctly from confinement.

• The “de-confinement” process should be handled carefully in order to avoid new outbreaks and to manage the medical and other critical resources.

The solution you bring to the table

• By means of a number of selected KPI (key performance indicators) based on data analytics and machine learning, we will provide real time status to help in the decision-making process during the different phases of the de-escalated process of the confinement.

• The KPI (key performance indicators) have been intentionally and conscientiously selected by the team to capture the pandemic evolution as well as the status of different parameters.

• By closely monitoring those indicators it is possible to have a correct follow up to contain and gradually solve the crisis. For instance, by tracking the reproduction rate, R0, which measures the average number of people who will contract a contagious disease from one person with that disease, we can evaluate if the spread of the virus is well contained by the measurements taken.

• The parameters (data) that feed the KPIs have been classified in clusters that can be easily modified, added or removed depending on the evolution or their suitability.

• The identified clusters are the following:

 Cluster 1: Epidemic data and statistics: number of cases, infected, ICU patients, deaths, recovered...),

 Cluster 2: Mapping of medical resources: number of hospitals, number of sanitary personnel (doctors, nurses), number of critical medical equipment (ventilators, ICU beds).

 Cluster 3: Personal protection equipment and other hygienic goods (masks, protective equipment for sanitary personnel, hydro-alcoholic solution), availability and stock of the required means.

 Cluster 4: Evolution of testing: number of people tested, kind of test (PCR or antibodies test...), and results of these tests.

 Cluster 5: Population mapping: Ages pyramid of the population, type of occupation (by tracking if tele-work/tele-education is possible or not), people habits, commuting.

Note: The software is modular, so the addition of new relevant clusters to gather information and extract new KPI will be easily done. Additionally, the set-up allows to extract the information/indicators by regions in order to help in the de-escalated confinement according to the status of the different places.

Our proposal is to control the de-confinement in a bottom up way, allowing gradual people’s movement in a free way but obviously by respecting social distance and the known “good-practices” (mask use, cleaning..etc):

  • initially around 3* km (or due to imperative reasons as today), *this value could slightly vary

    • then in current city/village - afterwards in the community/region
    • and being increased gradually depending on the evolution.

The application will warn the authorities, depending on the KPI evolution, anticipating if it is necessary to revert the measures and the people’s movement.

What you have done during the WE

• We have identified some of the different KPI (key performance indicators) we want to extract from the data basis in order to gather valuable information and parameters in a synthetic manner (R0, death ratio, % of available UCI beds/number of people, percentage of tested people...among others).

• We have identified different data sources for pandemic information, population and Geographical data from websites, however we need consolidated data from administrations/governments to fill our clusters.

• We have analyzed the technical complexity of the project, concluding that with consistent and enough data to fill our clusters it will be easy to extract the correct KPI which will provide guidelines to manage the de-confinement.

• We have worked on the business plan and the resources needed to develop the project. 1. We have analyzed the available means to develop the project. At CITIC, University of A Coruña, we have already statisticians and data scientists that can develop the project and provide expertise and advice identifying the KPI and the trends evolution.

The CITIC of University of A Coruña is a public institution that can provide the Hardware to host safely both the data and the application software, (See specific document in attachment, ref “Annex1”)

• We have identified the remaining to do. The activities to be completed are mainly linked to the data gathering and the development of the set of parameters to be tracked and added to the available prototype (see URL below). The data available in the prototype can feed mainly cluster 1 and some points for the other clusters. New data sets are required to get the complete mapping.

The solutions impact’s to the crisis

This project has a strong added value to solve the crisis and recover social and economic activities in a controlled way by reducing and managing the risk of contagion and the capability of the health&sanitary system to absorb the potential infected people until a vaccine will be available.

Additionally, our intention is to develop a powerful tool to learn from the current pandemic, helping us to understand better its evolution and to improve the management of future health threats.

The necessities in order to continue the project

 We need data from administrations/governments and from population surveys/polls to feed the proposed clusters.

 We need funds to support our teams from CITIC of University of A Coruña (at this stage we have identified a team around 4 or 5 people).

The value of your solution after the crisis

The idea is to gather a maximum number of data. The work and conclusions extracted from this project can be applied afterwards using the same principle (data analytics and machine learning) to help in the decision making of a new potential virus/ threat encountered in the future.

Available data can help to predict by machine learning how a new virus/health issue could evolve. It will help in the decision-making process to prevent developing a new pandemic and a new crisis.

By gathering information from the new virus/threat in the early stages it will be possible to predict the evolution based on data and evolution from this crisis. Also to understand if specific medical staff (such as ventilators or other) are required in order to anticipate if measurements should be taken before saturating the sanitary system.

The URL to the prototype

The work is already started with the current available data, refer to https://www.udc.es/en/covid-19/evolucion/ Current data comes only from Spain, the idea is to complete it with all European countries and/or world countries.

                                                                        Annex 1 

The CITIC * of University of A Coruña** is a public institution with Staticians and Data Management Scientists, soma of them are already involve in dealing with the Co-Vid Crisis. CITIC will provide us the equipment that we need to develope our tool. The CITIC main Data Center (CPD) has a hot-cold distribution layout, cooled via four APC InRows and a two-unit redundant cooling system. It has nine standard size racks and one extended size rack. The datacenter power load is protected by an APC Symmetra UPS (48KVA). It is equipped with, among other equipments: -Racks with servers and specific equipment, switches or switches for interconnection to the UDC fiber network, patch pannel. -Two IBM X3550 servers to provide services to the center.

  • Network storage system (SAN). IBM System Storage DS4700 Express Model 70 with switch FiberChannel IBM TotalStorage SAN16M-2.
  • Heterogeneous virtualization system. IBM BladeCenter H with different architectures (8 Intel Xeon blades, 4 PowerPC blades and 2 AMD Opteron blades).
  • SunFire T1000 8-station cluster with SPARC architecture for virtualization.
  • 6 ASUS ESC4000 G3 GPU servers with four Nvidia GTX 980 each.
  • 5 Lenovo nx360 (2 x Xeon E5-2650 v4, 128GB RAM), two of them equiped with a Nvidia Tesla M60. Configured in a VMware ESXi cluster.
  • 2 Lenovo nx360 (2 x Xeon E5-2650 v4, 128GB RAM), each one equiped with a Nvidia Testla K80
  • 2 Lenovo nx360 (2 x Xeon E5-2650 v4, 128GB RAM), each one equiped with a Nvidia Testla P100
  • Storage Server for the VMware cluster installed on a Lenovo X3650 (2 x Xeon E52650 v4, 128GB RAM, 1TB HDD for OS), with 3.5TB of SSD and 20TB HDD.
  • Cluster for high-performance computing (HPC) of 16 multicore nodes, GPUs and hardware accelerators, interconnected by InfiniBand network.

The Centre for Information and Communications Technology Research (CITIC) is one of the four research centres of the University of A Coruña, which main objective is to promote the advancement and excellence in research, development and innovation in Information and Communications Technology (ICT) and to promote the transfer of knowledge and results to the society. *The University of A Coruña (UDC) is a public institution founded in 1989 whose primary objective is the generation, management and dissemination of culture and scientific, technological and professional knowledge through the development of research and teaching.

Built With

  • caret
  • dataquest
  • h2o
  • infrastructure/equipament-of-citic-udc
  • r
  • shyny
Share this project:

Updates