IEEE Circuits and Systems Magazine - Q3 2021 - 73

Team: Sebastián Marzetti1, Pierre-Alexandre Peyronnet2,
and Florent Barthélemy2
Mentors: Valentin Gies1,2, Valentin Barchasz1, Thomas Delaey-Langlois3,
Daniel Peloux2, Philippe Arlotto2, and Hervé Barthélemy1
1Université de Toulon, Aix Marseille Univ, CNRS, IM2NP, Toulon, France
2Université de Toulon, IUT de Toulon, France
3Direction Générale de l'Armement (DGA), France
Corresponding author: V. Gies (email: vgies@univ-tln.fr).
Low Cost Artificial Ventilator Embedding Unsupervised
Learning for Hardware Failure Detection
Abstract
In this paper, a less than $200 artificial ventilator that can be used
against COVID-19 pandemic is presented. Using low-cost easyto-find
materials,
it has been designed for helping developing
countries where supplies for building new medical equipments
are limited.
It complies with medical requirements, allowing to
monitor and adjust ventilation parameters such as tidal volume,
maximum intra-lung pressure and breath rate.
Even if this ventilator is low cost, focus has been placed on
improving its global reliability. Using low-cost recycled materials
may lead to mechanical failures, this potential drawback is
addressed with an intelligent embedded hardware failure detector
implemented inside the microcontroller. Using K-means
optimized algorithm, it learns in a short time normal operation
corresponding to the couple formed by a given ventilator set-up
and a patient. In case of a mechanical breakdown, an alert is
generated to inform medical staff.
First, mechanical, electrical and software architectures of the
system are presented, then hardware failure detection algorithm
is detailed. Finally, test results done at IRBA using an artificial
lung are discussed. The overall project has been published as
an open source one on GitHub: https://github.com/iutgeiitoulon/
ArtificialVentilator.
I. Introduction
D
uring COVID-19 initial lockdown period, a low
cost (less than $200) artificial ventilator has been
designed at Toulon University, in France. It implements
fully autonomous ventilation mode for reanimation,
and assisted ventilation mode where patient breath
start is detected and assisted by the ventilator. It complies
with artificial ventilator medical requirements: limitation
of inspired tidal volume, intra-lung pressure and
adjustable breath rate [1]-[3].
Using easy-to-find low-cost materials such as fishing
wire, aquarium tubing, hose and 3D printer stepper
motors, this artificial ventilator can be built anywhere
in the world through standard tools such as a jigsaw or
a laser cutter. This makes it useful in emergency situDigital
Object Identifier 10.1109/MCAS.2021.3092539
Date of current version: 12 August 2021
THIRD QUARTER 2021
II. Disclaimer
This artificial ventilator is not intended to become a
commercial product used in a modern hospital. It has
not been certified as compliant to state of art medical
regulations and will not be asked for. However, it has
been functionally tested successfully on an artificial
lung as shown on Fig. 1 at IRBA (Institut de Recherche
Biomédicale des Armées) in HIA (Hopital Inter-Armées)
Sainte-Anne, Toulon, France. This test has been detailed
in written report. However, Toulon University students
and researchers cannot be liable of any damage, injury
or death resulting from the use of this device.
III. Mechanical Architecture
Proposed system mechanical architecture is presented
in Fig. 2. It can be built using only material plates such as
wood or plastic, using a jigsaw or a laser cutter.
IEEE CIRCUITS AND SYSTEMS MAGAZINE
73
ations where it is not possible to have an access to a
medically certified ventilator, such as during COVID-19
pandemic. However, quality of these materials is unknown.
Therefore, in order to improve robustness, a
solution based on cooperation between embedded artificial
intelligence and medical staff has been developed.
It aims at warning medical staff when hardware failures
are detected using unsupervised learning algorithms.
In order to help people, Toulon University students
and researchers have decided to open source this
project. Electronics, mechanics as well as embedded
and monitoring software have been shared on GitHub:
https://github.com/iutgeiitoulon/ArtificialVentilator.
They can be used or modified freely.
The article is divided in five parts. Sections III and IV
present ventilator mechanics, electronics and embedded
software architecture. Section V presents monitoring
software structure. Then section VI focuses on how
to improve system robustness with an embedded failure
detector using K-Means optimized algorithm. Finally, in
section VII, results are presented and discussed.
https://www.github.com/iutgeiitoulon/ArtificialVentilator https://github.com/iutgeiitoulon/ArtificialVentilator https://github.com/iutgeiitoulon/ArtificialVentilator
IEEE Circuits and Systems Magazine - Q3 2021

Table of Contents for the Digital Edition of IEEE Circuits and Systems Magazine - Q3 2021
