Voir en


Introducing CARA, CERN’s COVID Airborne Risk Assessment tool

CARA, CERN-made software modelling the concentration of viruses in enclosed spaces, will inform space-management decisions on the CERN sites


CARA graphs
The different graphs represent the concentration profile of viruses in air over the course of a working day, including a one-hour lunch break. The relative probability of infection can be visualised by comparing the area below each graph: the larger the area, the higher the chances of being infected. (Image: CERN)

Since the onset of the COVID-19 pandemic, CERN has favoured a proactive response to the virus outbreak by putting in place strict COVID-related health and safety measures. These include instructions concerning the use of workspaces and meeting rooms with the aim of minimising the risk of airborne transmission. The most recent outcome of this dynamic approach is the release of CARA, CERN’s COVID Airborne Risk Assessment tool, which uses and promotes scientific understanding of the risk of airborne transmission in indoor spaces at CERN.

Initially developed by Andre Henriques of the Occupational Health & Safety and Environmental Protection unit (HSE), the tool is receiving valuable additional contributions from engineers and physicists in the Experimental Physics (EP), Beams (BE) and Information Technology (IT) departments (full author list). It models the concentration profile of potential infectious viruses in enclosed spaces with clear and intuitive graphs. The user can set a number of parameters, including room volume, exposure time, activity type, mask-wearing and ventilation. The report generated indicates how to avoid exceeding critical concentrations and chains of airborne transmission in spaces such as individual offices, meeting rooms and labs.

Consequently, CARA enables space managers and safety officers to judge whether or not the baseline measures in place deliver an acceptable risk level and are adapted to the workplace. By comparing how virus concentrations in spaces vary with factors such as closed or open windows or HEPA filtration systems, the software gives clues as to which measures should be adopted to minimise the risk of transmission and can guide targeted investments in technical solutions such as ventilation or filtration systems.

“The original idea of CARA was to investigate scientifically the risks associated with, for example, a speaker removing their mask during an event in the Council Chamber or people sharing offices,” explains Andre Henriques.

The original model was coded using Wolfram Mathematica. Two different versions of the CARA tool are now available in Python with a web interface for anyone at CERN to use. In the first, the user completes a simple web form and the calculator app generates a quick assessment of virus concentration in different workplaces. The expert app has a more detailed parameter interface for safety officers, space managers or department-level supervisors to tailor measures to the areas that they supervise.

While the software is already accessible and provides accurate results on potentially inhaled doses of viruses in given settings, future upgrades will better reflect some parameter uncertainties for an even more detailed and precise result. Originally meant for internal use only, the software has knowledge-transfer potential for medical applications as well: together with the CERN Knowledge Transfer group, the CARA team has started collaborating with the Institute of Global Health at the University of Geneva and the BioDynaMo project.