Blue Haiven: Covid-19 Response Options for UC Berkeley

Proactive COVID-19 Response Options

University of California, Berkeley

______________________________________________________________________________

Table of Contents:

1. Introduction

2. Smartphone and Smartwatch Applications

3. Overhead Monitoring System

4. Building Protocol

5. Road Map

______________________________________________________________________________

Overview

This guide is designed to serve as a road map for ways that UC Berkeley can be proactive in their efforts to prevent widespread outbreaks while being able to continue essential campus operations. The plan will cover each individual component as well as how they will work in synergy.

The critical idea behind this plan is being able to detect when someone’s immune system is responding to an infiltration. Then, the system must be able to determine who they came in contact with during a potentially infectious period and send out proper alerts in order to prevent further spread.

It should be noted that the goal is not to completely open the campus, but rather to allow small groups to begin operating at close to normal levels. As a suggestion, this should be implemented by research labs first, as they have small numbers of people that consistently interact with each other. Then, other operations may start opening up as long as there are no prevalent outbreaks. Buildings with circulated air, such as libraries, should not be accessible for long periods of consecutive time.

The contents of this plan should be used in combination with preventative measures suggested by other authorities, such as social distancing both within and outside of buildings, as well as required usage of masks.

______________________________________________________________________________

Authored by Devin Cox, B.A. Data Science

For questions, please reach out to devincox99@berkeley.edu

Edited on May 19, 2020

1. Introduction

Key Concepts:

1. Contact-Tracing

   1. This will be implemented by using low-energy bluetooth on smartphones as well as using bluetooth beacons that can be installed within buildings to provide high accuracy on detecting interactions between two people.

   2. All interactions will be recorded locally on each smartphone, recording the bluetooth device ID, a timestamp, and a location.

2. Immune System Response Detection

   1. This will involve using the combined efforts of monitoring heart rate, oxygen levels, and skin temperature in order to detect an illness before symptoms arise to a noticeable level.

   2. By utilizing this method, the system can bypass requirements for user input, however it can still be supplemented by user’s answering a daily survey about their current conditions.

3. Efficient Alerts System

   1. The server must be capable of analyzing in real-time. When an infection is discovered, each smartphone will search locally if its user was potentially infected (based on bluetooth ID’s and timestamps). If a match is discovered, that user is notified and their own ID is marked as potentially in danger of illness.

4. Overhead Monitoring System

   1. School officials may have access to understanding student and faculty health on a broad level so that they may make an informed decision for immediate shutdown if necessary.

   2. An individual's ID is connected to their health status, which can be utilized to enforce usage of the protocols listed later on.

Equipment:

1. Smartwatches (1 per person)

2. Smartphones (1 per person)

   1. Suggestion: Fitbit Versa 2 or Apple Watch

   2. Key responsibility is to monitor heart rate in real time

3. Pulse Oximeters (Either 1 per building or 1 per person)

4. No-Contact Infrared Thermometers (1 per building)

5. Bluetooth Tracking Beacons (1 per room and/or 30 square meters)

1. Smartphone and Smartwatch Application

The following is a rough outline of important considerations for the applications.

1. Local Contact Tracing

   1. Each smartphone stores its own data, therefore relinquishing the need for a large database to be managed. When someone becomes sick, every application is updated with an ID, and internally computes whether or not an interaction took place.

   2. Centralizing all the data could potentially be more useful, but also likely more costly.

2. Privacy Concerns: Data Dissemination and Randomization

   1. It will be important that everyone is kept anonymous from each other, and that a person’s identity is only retrievable in necessary circumstances.

   2. To an extent, those involving themselves with this program will be foregoing aspects of their privacy.

3. Daily Surveys

   1. These surveys must be completed before a person enters a facility.

   2. The format will be a short series of questions, asking about potential symptoms (coughs, quality of sleep, etc.) A person will grade it on a 1 to 10 scale, and there will be a place for additional comments if they wish to do so.

4. Alerts and Notifications System

   1. Alerts user if their health seems to be declining, requesting that they shelter in place for a specified number of days.

   2. Alerts user if they have interacted with someone over a potentially infectious period, and requests that they shelter in place.

   3. Allows for persons with special permissions to send out a statement via the app, such as an immediate shutdown of campus operations.

5. Heart Rate Monitoring

   1. Consistently track current and average heart rate.

   2. Determine baseline after short period (1 to 2 days).

6. Sleep Cycle Monitoring

   1. Use in combination with daily surveys to help provide a bigger picture of user well-being.

7. Image Recognition System

   1. Ability to take pictures of oximeter and infrared thermometer after usage and automatically update internal data with results.

8. Local Anomaly Detection

   1. The computation does not look to diagnose, but rather determine when something “unusual” is happening, including low oxygen levels or an extended period of elevated heart rate.

1. Overhead Monitoring System

1. Centralized Access to Data

   1. Provides easy ability to do further analysis if deemed necessary.

   2. Allows executives to make informed decisions on next steps.

2. Notification System

   1. Centralized and only specific users have permission to send out alerts to everyone.

3. Application Communication

   1. Ability to take in an alert when someone gets sick, and then notify all applications of the ID so that they make necessary internal computations.

4. Personal Checks

   1. If necessary, check a specific individual's status (including a “validity” check that looks to see if they have been participating in the program to its complete extent).

1. Building Protocol

1. Frequent Oximeter Checks

   1. Should occur at least once a day and before anyone enters a building.

   2. Can make immediate decisions if someone’s oximeter reading is too low.

   3. Ensure proper sanitation after each usage.

2. No-Contact Infrared Thermometers

   1. Similar usage to oximeters, however not as effective for preventative measures.

3. Limitations for Buildings with Circulated Airflows

   1. Circulated airflow can cause the virus to spread easily regardless of precautions.

4. ID-Checks for Enforcement of Protocol

   1. If deemed necessary, a person’s “validity” (check if they have been consistently using applications) can be connected to their student or faculty ID and scanned by building personnel.

5. Mask Requirements and Social Distance Practices

6. Bluetooth Beacons

   1. Beacons should be placed every 30 square meters for precise contact tracing when indoors.

1. Road Map

Stage 1: Build software programs

Stage 2: Install and distribute equipment

Stage 3: Collect baselines

Stage 4: Start operations