The virus is an invisible enemy. Anyone can be the source of other people's infection. We need to cut that chain of spread by making the virus visible and prepare ourselves.

THE SCENT OF COVID19 wants to detect the presence of the virus through the analysis of people's breath.


We will defeat the virus by making it visible. While the Covid-19 widespread is plateauing in some countries, governments are actively looking for solutions for the so-called "phase 2" to lower the lockdown. This is a critical moment, where measures to avoid a new burst are necessary. In light of this, we have concentrated our effort into designing a solution to detect the COVID19 through a sensor that is accessible and easy to deploy.

What it does

The aim of the Project is to define the breathprint characterizing COVID19 at different stages of the pathology (healthy people VS without clear symptoms VS with symptoms) in order to use the data collected for the development of a fast, reliable and sensitive diagnostic method.
This aim can be achieved with the analysis of VOCs in people's breath through a DIY sensor.

Practically it would be accomplished through:

  1. A low-cost DIY kit consisting of a VOC sensor array, a controller card and a mobile application.
  2. A platform to share this kit, with financial resources so that companies can build it themselves and provide to the distribution.
  3. An open database that concentrates all these data, anonymized and standardized.
  4. A neural network that analyses these data, contrasting the VOCs of the breath and your health status, to determine the respiratory footprint of the COVID19 at different stages of the infection.

With this kind of information it will be possible to use the DIY as a diagnostic instrument: according to its breathprint, in fact each person performing the test will be assigned to a predefined category.
In this way

It will be possible to rapidly distinguish healthy subjects from asymptomatic ones.

How we build it

The biomarker of COVID19 in respiratory air is not known.

Through a campaign of testing of people with and without the symptoms of the pathology, carried out through the Kit that detects VOCs from expired air, it is possible to generate an open database. Analyzed through Artificial Intelligence algorithms, perfected through the KAGGLE platform, the aim is to search for the virus' footprint at different infective stages.
This will lead to the determination of the parameters that will be used for non-invasive diagnosis of the population.


  • Identify the VOCs (volatile organic compounds) that the body generates when it is infected with COVID19
  • Develop a device capable of detecting this VOC
  • Develop the DIY kit (array of VOC sensors)
  • Find volunteers who want to make the kit, test it and send the data
  • Generate a quick iteration of the solution -where needed- according to initial test results
  • Finding a means of financing, to facilitate the manufacture of the kits
  • Develop an AI algorithm that finds the virus' VOC breathprint in exhaled air
  • Train the AI model as quickly as possible in order to start generating reliable predictions.

    Looking at the economic crisis caused by the coronavirus, a business model that allows a way of constraining the price of the solution is needed so that people without the means to buy the DIY kit could have access to it. Need for quick and reliable distribution.

Accomplishments that I'm proud of THE SCENT OF COVID19

Achieving a fast, cheap and simple methodology that can detect the virus in asymptomatic patients, to anyone and anywhere.
Being able to build an effective approach that is accessible, as well as easy to distribute, making it available to as much people as possible so that we can make a positive difference in this Covid19 crisis.

What I learned

  • We have seen the amount of dispersed and disconnected information about the COVID.
  • We have learned about VOC biomarkers from people's breathing air and the sensors to detect them.
  • That the only way to know and beat this virus is through collaborative knowledge, the one we have had and the volunteers who test the kit and the optimizers of the algorithm.

What's next for

  • Developing the kit
  • Testing
  • Building partnerships and implementing a commercialisation strategy:
    • Via funding or sponsorship of the DIY kits with public authorities as well as private organisations (which are actively looking to collaborate in CSR plans).
    • Via crowdfunding: buy one-donate one campaign
    • Via partnership: similar to sponsorship but in this case joining forces with hardware providers to launch a constraint price quickly
  • Dissemination
  • Developing marketing channels: website, social media...
  • Connecting with the +800 NGOs, foundations, associations, and civil society groups, universities and public authorities linked to the panEU hackathon
  • Distribution kickoff
Share this project: