On this website we use cookies that are strictly necessary to make our website function correctly, as well as optional – analytics, performance and/or marketing - cookies which help us improve our website by collecting and reporting information on how the website is used, as well as help us to reach out to you with information about our organization or offer. If you do not wish to accept optional cookies when visiting our website, you may change the settings in the Cookie Settings. For more information go to Cookie Settings.

Skip to content

Our Response to COVID-19: A Fully Automated Health Check Station

Technology

May 27, 2020 - 4 minutes read

v01_HackCovid_19_Blog_768x440
Objectivity Innovative leader in technologies

Our specialty is designing, delivering, and supporting IT solutions to help our clients succeed. We have an ethical framework that underpins everything we do. Our underlying philosophy is that every client engagement should result in a Win-Win and this is supported by our four values: People, Integrity, Excellence, and Agility. Our clients are at the heart of our business and we are proud to form long-lasting working relationships, the longest of which is 29 years. Our goal is to continue to grow our business whilst remaining true to the ethical framework and values on which we are founded.

See all Objectivity's posts

Share

We recently ran a hackathon in our AI team the goal of which was only broadly defined as – “What should be our response to COVID-19?”. The hackathon teams approached the problem from many different angles. This article describes what one of our AI teams has managed to accomplish, why we did it, and how.

The Goal

When COVID-19 restrictions start being lifted, people will gradually return to their offices; (for some of us introverted IT people, this may not necessarily be only a good experience 😊). But, even when the restrictions are lifted, this doesn’t mean that everyone will be healthy from then on. People will get sick, and sometimes they may not even realise it. So, in light of this, there should be some kind of health check at the office reception to verify employees’ state of health and which would signalise when certain people could potentially be sick. Ideally, we would need someone to scan every person that comes in – someone who would be able to react appropriately, if necessary. Unfortunately, not every office can afford to hire a full-time nurse. Plus, this would create additional risk, as this person would be highly exposed to contact with many people. Thus, the idea for our response to COVID-19 was born – let’s create a fully automated health check station!

Ridiculous Constraints

No good hackathon can take place without applying certain ridiculous constraints. This was also true this time around, as: - We devoted only 2 days to hacking, - The team consisted of only 3 people, (including 1 manager which takes it down to 2, at best), - Due to isolation, we had to work in separation, meeting only for 2 hours in the final stage of testing, - We didn’t want to buy any expensive devices – instead, we wanted to construct our health check station from what we already had at home.

The Purpose

Our use case was simple. An employee enters the office lobby. They approach the health check station. They scan their office card to authenticate. Next, they move their forehead close to the scanners. The scans are done automatically. The result of the scan is showed on the station’s display. If the scan results in a potential health risk situation, the following scenarios are executed: - The scanned person is informed about the scan result and the steps they should follow, - Appropriate individuals in the office are informed about who was scanned and what their result is in order to be able to take certain necessary action, i.e. quarantine co-workers, etc. The health station is then ready for the next scan. Do you want to see our prototype in action? Have a look at this video:

How Does It Work?

Some say that a picture is worth a thousand words:

Hackathon COVID-19 temperature measurement app diagram

Still, our picture requires at least some words 😊. From the top, we have a laptop with a regular camera, which is our main processing unit and display. It’s connected to Raspberry Pi through WIFi which has a working card scanner (pigpio). It also has a distance sensor (to recognise that the user’s forehead is close) and a servo mechanism connected to it. In the prototype however, due to the team members separation and varying toolset, we used both Raspberry Pi (card scanner) and Arduino (distance and servo). The servo triggers a regular hand thermometer a few times in a row. The camera records the readings, transforms them (OpenCV + wx) and sends them to the interpretation module. To interpret the readings, we tried to use various OCRs, but without any luck. So, we trained a neural network (Keras + TensorFlow) to recognise images – this worked out great. Finally, the laptop shows the scan result to the user on its display.

How Much Does It Cost?

If you would like to build one for yourself, the cost of the hardware you’ll need can turn out to be quite minimal:

- An old laptop with a camera

- RaspberryPi with WiFi

- starting from $15

- RFID card reader - $5

- Servo motor SG90 - $1

- Ultrasonic Distance Sensor HC-SR04 - $1

- Micro USB power supply

- Touchless thermometer

- starting from $60

- Some LEDs and a beeper if you want the device to signal in this way.

Let us know if you would like to create your own automated health check station – we can help!

How Could This Be Further Developed?

You can’t build much during a 2-day hackathon, so you need to limit the scope to what you can deliver in this kind of timeframe. But, our vision was bigger – we were also looking to equip our health check station with:

- Age and gender estimation (e.g. using MS Cognitive Services),

- Pulse measurement,

- Breathing sound analysis (yes, this could be hard, but there is some research around this already).

Summary

In a very short time, with limited people and in a difficult working environment, the team achieved remarkable results 😊. With the help of a lot of cardboard, cable ties, sensors, and AI, we managed to construct a working prototype of an automated health check station. Our response to COVID-19.

Get in touch if you’re interested in building an automated health check station yourself. 😊 Stay home, stay healthy!

About the Authors

Marcin Faber Data Scientist

Data Scientist who used to be a Graphic Designer and Frontend Developer. Loves projects where combining knowledge and data from various fields with creative and out-of-the-box thinking is required. Will not despise the custom visualisation of data or process flows. In his spare time, he enjoys participating in side projects and outdoor adventures.

Jacek Krasnoborski Data Scientist

Data Scientist at Objectivity with a background in software development. His main field of interest are topics related to Deep Learning applications for Computer Vision problems. Coffee geek and climbing enthusiast.

Michał Zgrzywa Data Science Guild Master

AI practice leader at Objectivity, experienced manager, software developer at heart.

Objectivity Innovative leader in technologies

Our specialty is designing, delivering, and supporting IT solutions to help our clients succeed. We have an ethical framework that underpins everything we do. Our underlying philosophy is that every client engagement should result in a Win-Win and this is supported by our four values: People, Integrity, Excellence, and Agility. Our clients are at the heart of our business and we are proud to form long-lasting working relationships, the longest of which is 29 years. Our goal is to continue to grow our business whilst remaining true to the ethical framework and values on which we are founded.

See all Objectivity's posts

Related posts

You might be also interested in

Contact

Start your project with Objectivity

CTA Pattern - Contact - Middle