Intuitively SEnsing COronavirus Risk
DEVICE OVERVIEW
What is this?
I wanted to create a way for people to intuitively feel ventilation and COVID-19 risk. By using my CO2 sensor wristband and the Buzz, users are able to feel the approximate amount of ventilation in a space and make informed decisions based on that information.
Since coronavirus is spread via respiratory droplets, a major factor in risk is ventilation. Total amount of exposure to the virus is what seems to ultimately decide if a person will be infected and how severe their symptoms will be. This means that it matters how much time you spend in a space. Ventilation and total viral exposure are pretty abstract to reason about. There are too many variables: time, air circulation and size of the space that are difficult to understand intuitively.
How Do I use This?
Step 1 - Pair the Buzz. You may need to reflash the program on the Adafruit Feather so that it is able to detect and connect with the Buzz.
Step 2- Assemble the sensor wristwatch. Insert battery JST into the Feather connector. Place the battery-Feather assembly into the wristband holder. It should slightly press fit onto the pins.
Step 3 - Do some sanity check testing by blowing air onto the sensor and feeling the Buzz. The vibration intensity and frequency should increase with CO2 levels and the numbers of motors actuated should increase over time. The sensor is not very reactive and it may take up to 5 seconds for the Buzz to respond.
Step 4 - Go out in the world! Go into different places and shops and restaurants and feel both safe and horrified at the different places you frequent.
Build
Parts List:
Female and male header pins
ON sensor selection
I selected the SCD30 module primarily because of its internal compensation and long-term accuracy. eCO2 sensors are cheaper and more widely available, but need to be calibrated against known quantities.
ON MEchanical design
This is a simple box to hold the Feather that was designed in Fusion360 and printed on a Prusa 3D printer. The battery goes underneath Feather and the CO2 sensor goes on top. The electronics are open to maximize airflow around the CO2 sensor and held together with header pins.
BLOCK DIAGRAM
Thoughts
Expectations vs. Reality
When I first went into this project, I assumed I would be disappointed at how many different businesses and buildings were not following re-opening guidelines. I was actually pleasantly surprised at how many places were actually doing a fairly good job.
Every week I go to the hospital to get immunotherapy for allergies and I was worried that the hospital wasn't doing enough to make the spaces safe. Most of the time there was no discernible airflow. When walking through the halls or past people in the waiting room, I would hold my breath. However, for the majority of my time, the CO2 remained fairly close to 400 ppm - there was little to no motor actuation.
In some ways, this device had the opposite effect for the places I surveyed. Instead of filling me with dread, it has made me feel safer in some of the places I am required to visit. Of course, hospitals are still a fairly risky place, regardless of ventilation, it has made me more informed and intentional about risk.
Airflow vs. Ventilation
I observed the difference between airflow and ventilation while doing the car experiments. After raising the CO2 levels in the car, I set the AC to medium fan strength. Despite significant airflow, the CO2 levels in the car barely declined over a 10 minute period.
This makes sense, given the example of the Chinese restaurant that had an outbreak of COVID-19 associated with air conditioning.The restaurant, like the car, had good airflow, but poor ventilation and was able to facilitate droplet transmission instead of preventing it.
This highlights why intuitively understanding COVID-19 risk is so difficult. Humans can only sense airflow - we can't sense ventilation at all. Somewhere with strong airflow may seem safe, but still be incredibly risky and there is no way for us to tell.
Appendix A - FAQ / mini Literature Review
Is ventilation actually an important factor in COVID-19 spread?
There have been studies correlating increased ventilation with decreased infectious disease. This is a review paper on different methods of trying to better understand the relationship between ventilation and infectious disease spread. There is a clear inverse relationship, but there is a need for more quantitive and experimental studies to fully characterize.
This paper in Nature ran simulations on aerosol spread based on position. In their research they determined: "In the pure aerosol-based model, bringing all rooms to the recommended ventilation rate would almost completely eliminate the chance of an outbreak". Another paper presents a hypothesis on why HVAC is an important disease control measure.
Although it is not scientific literature, I also like the Atlantic article We Need to Talk About Ventilation which gives a good overview of ventilation as it relates to COVID-19.
Is CO2 actually a good proxy for ventilation?
Generally, yes. It has been used as a "natural tracer gas for estimating the [ventiliation rate]" in schools and in HVAC systems. The sensor I'm using is designed for use inside an HVAC system.
This paper titled Review and Extension of CO2-Based Methods to Determine Ventilation Rates with Application to School Classrooms reviews different methods to estimate ventilation rate from CO2 readings.
Appendix B- Data <WIP>
I collected a fair amount of data during the testing of this device. Although a dual sending and receiving mode is available for the Adafruit nRF52840 Feather, I collected some of these data without the Buzz actuation because there was lag from running as both that I was not able to fully resolve.