Ventilator based on the MIT OpenSource project

"Respire" Emergency Ventilator Project

Project Timeline

April 3rd 2020

We are currently working on a Emergency Ventilator based on the Open Source project E-Vent from MIT


This project automatic patient ventilation by motorizing standard manual ventilation bags (Ambu) which every hospital uses. It’s a simple and inexpensive solution to make.

We took the CAD files in order to adapt them to the metric system (the files provided by the MIT are in inch, difficult to realize in Europe because we can’t find the components : plate thickness, diameters, etc…).

The aim is to be able to produce the parts by laser/waterjet cutting, 3D printer and on a standard router, without any other tools.
E-Vent Metric

CAD files in STEP format can be downloaded free of charge here :

April 13th 2020 : Version 0.80 here : http://www.qualup.com/_download/e-vent-metric-v0.8.zip
(same Creative Common CC BY-SA 4.0 license as the MIT files)

These files are without any guarantee of operation and will be used to make the prototype .

The E-Vent project has the advantage of using a Arduino Mega quite easy to get for less than 35€ and inexpensive standard basic components.

Many thanks to the volunteers at MIT for having made available this project so clever and astute.

We’re going to do a bill of materials for the mechanical components as soon as I get online. We have made some technical choices dictated by the elements we have in stock (for example Rivpress nuts), you can adapt the files.

Our developer is currently creating a nomenclature of electronic components and potential suppliers in France-Europe that we will put online as soon as it is completed.

The technology used is quite close to that of a 3D printer, that’s why we are launching a prototype which, hopefully, should run next Friday, April 10th.
I’ll post the progress regularly, because it’s sure that we’re going to encounter some problems and unforeseen events.

All our subcontractors participate in the project., a lot of them provide prototype parts free of charge and some that were closed have reopened to help us.

Important :

The head of resuscitation at hospital in Mâcon is ready to test the prototype on mannequins during the development and thanks all the people involved in this project.

This project is not a Respirator, but a Emergency Respirator, to be used by the hospital services in the emergency cases who will decide it and under their responsibility.

The device presented is not a medical device, it is not homologated, not certified by the Regional Health Agencies, not certified by the French Ministry of Solidarity and Health nor by a any organization.

It was carried out for the sole purpose of compensating for an emergency situation where the equipment would run out and cannot replace certified equipment in terms of performance and reliability.

The sole purpose of this project is to help caregivers save lives.

Some info on the mechanics:

Sheet metal : 2017 (Au4g) 3, 4 et 6mm

Motor : Crouzet  Ref. : 8983b1, 12V CC, 15Nm, 30tr/mn (data that we found on the site e-vent.mit.edu, After a day of testing with Loic from Crouzet, it is correctly dimensioned)

Gears  : module 1.5 laser or water jet cut

Screws M4,  M5 et M6  Standard (Cergy-vis ou autre)

E-vent Metric 02

Patient ventilation parameters

(translated from MIT by Deepl.com)

Air volume (TV) : The total volume of air to be delivered to the patient.

BPM : Breaths per minute, also called respiratory rate (RR). It usually varies between 8 and 30 BPM.

Ratio IE (IE) : The relationship between the duration of inspiration and the duration of expiration. For example, a ratio of 1:3 means that the exhalation phase lasts three times longer than the inhalation phase. It generally varies between 1:1 and 1:3, with a maximum of 1:4 currently observed in COVID-19 patients.

In addition to these clinician-defined inputs, the high-level controller uses two other inputs: motor encoder position and system pressure. Its job is then to translate all of these inputs into the motor commands the low level controller needs: desired motor speed and position.

Note that since the unit does not directly measure volume, the tidal volume (VT) input of our controller is specified as a percentage of a full compression of the Ambu bag instead of litres. The percentage (%) of bag compression from 0 to 100% corresponds to the encoder pulses that correspond to the distance between the fingers of the device moving towards or away from each other and this determines the volume of air delivered.

Companies actively participating in the project :

Sinthylène – Pont de vaux 01 – Découpe, Usinages

ALPM – Mâcon 71 – Découpe Laser, Pliage

SpaceClaim Europe – Ville Sur Jarnioux 69 – Simulation Numérique ANSYS

Foyers Communautaires et Amicale Ancien Elèves La Source Lugny 71 – Financement Prototype

Solu-Watt  Mâcon 71 – Coffret Electrique

Onwi – Marcilly-d’Azergues 69 – Bureau d’Etude Electronique

ENSAM – Cluny 71 – Assistance Technique

+ A team of retirees uncompromising hyperactive…

All people / companies who want to help can contact me live: pboichut[at]qualup.com, I will do my best to centralize and dispatch . See my post of April 5th below explaining our needs.
Unfortunately given the number of candidates, we will have to make choices which will be arbitrary given the urgency, if you are not selected, it is certainly not due to your skills, but to other factors (e.g. proximity), we apologize.
In any case I will post all our advances, problems, mistakes, successes… on this page every time there is something new.

If you wish to to help us financially, all donations, even small ones are also accepted, the final goal is to provide for free 15 ventilators to the hospital of Mâcon, contact me : pboichut[at]qualup.com, or at 03 85 33 29 35 (dial 2), thank you.


  • The material and information on this site is provided without any explicit or implicit guarantee.

No material on this site is intended to provide medical advice. All models are intended for experimental use only.



April 4th, 2020 : Update STEP files

Changes Made in CAD Files : E-vent metric v0.7.zip

  • Addition of 8.00 radius on the printed parts that compress the Ambu bag to prevent damage to the Ambu bag.
  • Adding the incremental encoder and its support
  • Correction of several design errors

Addition of the BOM of the Electronic Components of the Prototype.

E-Vent Metric V0.7
13 Avril 2020 : Version 0.80 ici : http://www.qualup.com/_download/e-vent-metric-v0.8.zip

(same Creative Common CC BY-SA 4.0 license as MIT files)

April 5ft, 2020

We urgently need help, would you like to participate in the project?

Although the MIT did a lot of work,making a prototype emergency ventilator in eight days is no small feat.
The goal is to produce in series as soon as possible after validation by the medical profession.
We need help, all people of good will who want to help us are welcome.

The following requirements are prioritized in order of time
(T) means that the need can be realized in telework, no need to be in our region (South Burgundy) to help us.
(TL) means that the need can be realized in telework and/or locally, but in the region Bourgogne Sud so that we can recover the work provided.
(TI) means that the telework is Impossible
, you must be in the region Mâconnaise/Chalonnaise to help us with possibility to go by your own means to Lugny – 71260.

Some needs require less than one hour of work per day.

When the line is crossed out, displayed in green, and a First Name is added at the end of the line, it means that we have found the person(s), but contact us anyway if you think you can help us, our needs evolve daily.

The needs URGENT are for the week from April 6th to 10th, for the SERIES the needs are from April 8th to 30th (provided that we have succeeded in making the prototype)
1 – (TL) 1 Developer ArduinoKevin
2 – (TL) 1 Experienced Electronics Technician Richard
3 – (TL) 1 Turner-Miller (Traditional) Experienced – Jacques
4 – (TI) 1 Person to take care of the local logistics (picking up parts, bringing them to a company, looking for parts… ) STRATEGIC POSITION, this position can save us a lot of time – Requires a vehicle, gas, a phone, and a lot of availability between 8h00-20h00 in temporary schedule, Mâconnaise/Chalonnaise region indispensable. Roman
5 – (T) 1 Person who frequently monitors the e-vent.mit website. edu in order to inform us immediately of the availability of the source code of the software that manages the Emergency Respirator (search at least three times a day) + Analysis of the forum posts and search for any software source code on the e-vent.mit.edu. Loic site.
6 – (T) 1 English translator of web pages with knowledge of WordPress, Elementor and WPML, the goal is to take my daily posts and translate them into English independently (knowledge of WordPress + Elementor layout essential + WPML, know how to make a backup of the site before modification) Steeve
7 – (T)1 Person with skills in design office and/or methods: Production of nomenclatures from a STEP or ANSYS SpaceClaim file. Requires a license of a professional CAD software (SolidWorks, SpaceClaim or other…) able to generate good quality universal STEP files. Jean Louis and Jean Louis. Translated with www.DeepL.com/Translator (free version)
8 – (T) 1 Person with English and CAD skills to translate the component nomenclature and element names in the CAD file. Requires a license of a professional CAD software (SolidWorks, SpaceClaim…) able to generate good quality universal STEP files. Jean Louis and Jean Louis.
9 – (T) 1 Person with skills in social networks (Facebook – Linkedin) who can animate social networks in order to spread our needs by respecting the information in the paragraph “IMPORTANT” below. Florent
10 – Prototype Funding Need Community Homes and Friendly Former Students La Source
11 – (TL) Many people equipped with a 3D printer (FFF) who could print visors for us, we will need some of our machines for the Respirator prototype and have a lot of demand for the visors. Thanks for offering your services to the email below
12 – Test design: Writing a test protocol for the software that drives the Emergency Respirator Rémy (see with Kevin: kevin[at]qualup.com).

1 – 1 Manufacturer or distributor or importer of d.c. gearmotors who can help us : discounts on purchases, technical assistance (12VDC gearmotors, 30trmn, torque still in definition. …)Loic
2 – (TL) 1 lathe/milling machine capable of producing NC turned parts (very simple parts with large tolerances, not requiring a high end machine)
3 – (T) 1 Electronic/Electrical Engineer able to produce Industrial Documents: schematics, beam manufacturing plans, manufacturing ranges… from manual schematics, sketches, and photos of the protoBenjamin.
4 – (TL) 1 or more people with wiring knowledge: be able to carry out component procurement (preparation of component orders) and then carry out the electrical harnesses internally (South Burgundy).
5 – (T) 1 Person with documentation writing skills and knowledge of a documentation generation software (HelpNdoc, DrExplain or other…) Rémy.
6 – (T) 1 English translator of the different documents (otherwise a person able to use deepl.com and correct the translation errors of the automatic translator) Loic
7 – Financing of approvals of 15 Emergency Respirators after validation by the medical profession
8 – (TL) Locations to assemble the Respirators (LUGNY, MACON, TOURNUS – ~200m² separate rooms to comply with containment rules) Community Homes – Lugny
9 – (TL) Packaging, plastic bags, cardboard box…<

IMPORTANT if you want to help us, thank you!

But before contacting us, please read carefully the following :

– The goal is to provide for free as soon as possible Emergency Respirators to the hospital of Mâcon and at the same time to provide in Open Source all our documents and files necessary for the manufacture of such products so that others can benefit from them. We commit ourselves not to charge anything other than our purchases of raw materials, subcontracting and miscellaneous costs (postage, gasoline, etc…).

– The current context of Covid19 is totally changing the situation.
We are looking for experienced, autonomous, ready to work in a small team. Ready to work with common sense, ready to respect the team members, ready to be open-minded, ready to change their mind, ready to change their mind and span style=”color: #000000;”> to support me. The diploma is welcome but is not especially more important than experience or willingness to help. If you have no diploma but experience and skills in a field we are looking for, and want to help us, please contact us.

– We are a all small team with a limited budget and our time is limited, we may not be able to respond quickly to all requests.

– The list above requires some skills, if you don’t have these skills (which is quite normal), no need to contact us, just come back on this page regularly to see if our needs don’t change.

– If you have other ideas of Emergency Respirators, better done, but with other mechanical or electronic principles, Make the prototype yourself and/or look for another company.

No need to contact me by Facebook or Linkedin, I’m old school and don’t have time to look, the only way to contact us is by email: pboichut[at]qualup.com or mboichut[at]qualup.com. Translated with www.DeepL.com/Translator (free version)

Avril 6th, 2020

First mechanical tests of the prototype… We’ll receive the electronic components tomorrow.
Thanks to all those who offered to help us, in less than 12 hours all urgent tasks have been assigned.
Everybody knows the 80/20 rule, 80% of the work is done in 20% of the time, we have more than 20% left, quite a lot of work in perspective.

April 7th, 2020

Today, a hard day, we have received none of the products ordered, and no electronic components for the proto.
We spent our day looking for components here and there to be ready tomorrow, because an engineer from Crouzet (the engine manufacturer in Valencia) is coming to help us develop the prototype. The ENSAM in Cluny is also taking part in the development day.
Hopefully I’ll have better news tomorrow night.

April 8th 2020

After a big day with people supers, constructive and open (Richard : ENSAM Cluny, Loic : CROUZET and Kevin) With the help of the new software, we have almost solved all the technical problems of DC motor control.

The mechanics work well (well, it’s still only a proto), but we think we’re ready when the MIT will provide the sources of the control software .
Here’s a little video of the prototype with a bladder simulating a lung connected to the Emergency Respirator.

A HUGE Bug remains a ENORM… 😀

Tonight at 8:00 P.M. sharp, it would appear that she is aware of the problems the caretakers are having and that she has lost it. She got completely wrapped up, and started to plead with the caregivers to thank them for all the effort they’ve put into this difficult period ( although she isn’t even near a window !). We have analyzed the code and we ask ourselves why she is doing this!

We’ll monitor her closely every night around 8:00 p.m. to see if the bug happens again…
It’s not the April 1st but the 8th ! 🙂
No it’s a joke, it’s reliable… But after a 16h day, sometimes you want to relax.


April 9th, 2020

Another hard day“, but hey, the prototype is well advanced thanks to the company Solu-watt in Mâcon who did a great job! Imagine, I had received an email from Solu-watt telling me that they offered to help us. <Yesterday, I call them in the morning, and we discuss our needs… <Yesterday, in the late afternoon, in the late afternoon, strong, pascal, from Sol-watt, strong brings me a pre-wired box with all the fixtures, drills, screen, accessories… presentable’, because tomorrow afternoon, we’re going to the Réa of Mâcon to check if the prototype will correspond to their needs.

Another good news, the chain M6 is doing a report on the Réa de Mâcon tomorrow, and we will be able to present the prototype.

Thanks to Kevin who, once again, has worked countless hours to make the software work (since we launched this project on April 3rd, Kevin has been working voluntarily more than 12 hours a day, Saturday – Sunday included, impressing this youngster !)


PS: We think we’ve found the 8pm bug, she’s been very quiet tonight  😀

April 10th, 2020

Thanks to all of you, we managed to present the prototype in time, making such a prototype that combines: mechanics, electricity, electronics, software… In 7 days was not an easy task! I’m rather proud of the team that was mobilized around this project, thank you.

This afternoon, first presentation of the prototype to the person in charge of the Resuscitation of Mâcon.
We explained to him the different functions of the potentiometers, buttons and safety devices used.
He seemed satisfied and told us that the functions of vital safety were present.

The list of alarms of the device is here : e-vent-alarms.pdf

We did a test on a dummy, with a mask and then in intubation, everything went well.
The machine worked well on the dummy (we had never tried it on anything else than a Respirator calibration bladder).

Pfff… I was so focused on the project, that I didn’t even think about taking pictures, but M6 was there and filming the whole presentation, we’ll have to wait for the broadcast on Tuesday night.

The problem now is to know when MIT will make available the sources of the control software.
The key point is that without these sources, the machine doesn’t work !
We sent them several emails, but they don’t answer, we can understand why they are busy.

Seeing the fantastic mobilization that happened to make the prototype, we could develop this software ourselves.
But I think, even with the best will, we won’t release the software before MIT, which is almost 1 month ahead of us. I don’t see how we could be faster than them, it would be a total waste of energy and what’s more, no testing would be done.

A second team is starting in Valencia on this project, 2 developers will work in collaboration with Kevin to prepare and lock‘ all the routines specific to this machine (DC Motor Driver, Encoders, Pressure Sensor…).

The goal is to be ready to incorporate our routines in the MIT software when the MIT software will be available, to reduce the adaptation time and to be able to start the production as soon as possible.

A company joins the project: Onwi will make us the cards . More info this Week End.

April 11th, 2020

Here is the email received this afternoon from Doctor de Varax, in charge of resuscitation at the hospital of Mâcon, thank you to all those who participated in the project or who offered their help:

Cher Monsieur,

Merci de vous être déplacé hier en réanimation au centre hospitalier de Mâcon pour nous présenter votre prototype de respirateur particulièrement intéressant.

Je constate que vous avez énormément travaillé et au plus près des attentes de la réanimation, avec un modèle déjà validé sur ces bases, par l’équipe américaine de la MIT.

Nous sommes particulièrement touchés de toute l’énergie que vous mettez vous, vos équipes et toutes les sociétés qui vous accompagnent, dans ce projet tellement essentiel face à cette pandémie.

Nous avons ensemble regardé les quelques points sur lesquels nous devons travailler prochainement et qui seraient attendus afin d’optimiser la qualité et la sécurité de ce respirateur déjà très évolué.

Encore une fois un immense merci de la part de toute l’équipe de REA, des médecins et des soignants, et principalement au nom des patients qui sont porteurs de covid ou futurs porteurs et dont l’attente légitime se porte principalement sur la présence ou non d’un respirateur pour leur survie.

Veuillez recevoir, Monsieur, l’expression de ma profonde reconnaissance.

Docteur Roland de Varax

Chef de Service Réanimation

Président de la CME

Président de la conférence Régionale des Présidents

Centre Hospitalier de Mâcon

Bourgogne Franche Comté

We look forward to sources code from MIT.

April 12th, 2020

In our interview on Friday, Dr. de Varax told us that he would prefer, if possible, to have a one-piece ventilator because the space in the ABR is often limited.

Here is version 0.9 which integrates the electronics into the Emergency Respirator.

Hervé from Onwi, has made good progress on the electronic map of the E-VENT, to make the routing of such a map a Easter Sunday, hats off Mr Hervé! This will reduce the assembly and wiring times by several hours.

I got him on the ventilator in a matter of tens of minutes.

Event 0.9-01

It seems to me that it would be good to keep 2 versions, one with the separate control box (like the prototype) but composed only of standard, common components, and easy to supply (for example for developing countries), and one with the integrated control which is simpler to assemble but more difficult to supply, because the motherboard has to be made, which is not a problem in Europe.

If someone has experience and can advise us.


April 13th, 2020

STEP Files and list of electronic components of version 0.8 available for download here : http://www.qualup.com/_download/e-vent-metric-v0.8.zip (3Mb)


Improvements :

– Replaced the aluminium arms of 6 by an assembly of 3 stainless steel sheets of 2.0 assembled (the cutting quality is much better with 2.0).
– Addition of 6900-2RS bearings on the arm axles.
– Reduction of the play between the motor support plates and the front plate.
– Modification of the mounting of the arm axes to prevent the axes from rotating.
– Reduction of gears centre-to-centre backlash (due to improved cutting quality).
– Replacement of the original pickup sensor by a SS-10GL13 for better positioning accuracy.
– Modification of the pressing hands to have a better compression of the Ambu (we gain at least 15% compared to the original design). Note: The stainless steel arms are not modified but it works on our prototype and will allow evolutions of the hands, if needed.
– Modification of the supports of Ambu.

Nota :

1 – This design is designed for a motor Crouzet 8983b1, 12V DC, 15Nm, 30rpm without encoder, you will have to adapt the motor mounting according to the motor you are using (15Nm torque required)
2 – The encoder is temporary because we are going to use an engine with an encoder.
3 – Provide an Arduino Mega Card instead of the Arduino Uno in order to be able to integrate all security and alarms.
4 – Version 0.8 can be made if you plan to use an external electrical box, an arduino board and independent components (like our first prototype).
5 – Version 0.9 will use a sheet metal cover attached to the ventilator and a motherboard that can be provided at cost price to those who wish.

April 14th and 15th, 2020

The last 2 days were devoted to the creation of the bricks for the realization of the Emergency Ventilators.

  • The files of the laser cut parts are almost finished and we will launch tomorrow the laser cutting of the second prototype (with all the improvements) at ALPM, so that the team in Valencia can work as soon as possible.
  • Loic found us two Crouzet engines with encoders.
  • Hervé and his team are working on the motherboard.
  • Pascal worked on the wiring of the two V2 prototype boards.
  • Kevin and the Valencia development team are working on the software routines that will control the engine, so that they will be ready when MIT provides the sources. From what I’ve seen tonight, it’s progressed well.

Thank you all for your help.

April 16, 2020

We are now quite advanced to get a more accurate idea of the final cost of the Respirator.

Initially, the goal I was hoping to reach was a cost of less than 1000€ HT each.

The series should look like this:

Thanks to all the people who took part in this adventure, les good will who took part in this adventure, we’re there ! A respirator will cost less than 1000€ HT.

We have several contacts with different companies and/or organizations, I hope to have good news to announce soon.

All studies and software developments have been free and will be provided in OpenSource as soon as they are validated.

The biggest item is the motor (gearmotor 12V) : ~200€ HT. The electronic boards, another important item, will be manufactured by the company Onwi in the north of Lyon.

Several people asked me why, we did not use a wiper motor very cheap and with the same power. In fact it is for a reason of patient safety.
A wiper motor is not reversible and would prevent opening the arms in case of power failure or problem. Imagine that the system stops for some reason (power failure, electrical problem…) in closed position, it would be impossible to open the arms to remove the Ambu. You would have to disconnect the Ambu from the patient, find another Ambu and reconnect it. (Of course, it is always possible to design a manual mechanical decoupling, but this complicates the mechanics, and it seems preferable to propose a simple system that works well in an emergency, sufficiently tested and developed).

One thing that worries me a little. We’ve been contacted by l’ICAM (Institut Catholique des Arts et Métiers) from Kinshasa (Democratic Republic of Congo), who want to make Respirators

After talking with one of the people in charge of the project, I have the feeling that they are not at the end of their sentence! No laser cutting, very few machines (two entry-level 3d printers, welding station, column drill and disc machine…).

We gave them everything we could, but unfortunately, I don’t see how they could realize their project with what they have as tools.

Here is their 1st Prototype, full of tricks, made with a recycling engine.

Prototype ICAM Kinshasa

They’ll probably need a lot more help than we do…

We’re thinking about a way to contribute to their project.

April 17, 2020

1 – It’s getting better…

More exciting news about the COVID19 pandemic; the news reports that patients are coming out of the ICU, the hospitals are no longer saturated, resuscitation beds are being released… Despite the fear of a new wave with deconfinement, a spike and it’s good for all the caregivers on deck for so many days.

So you tell me: “Your ventilator project is too late”?  Well, so much the better if it won’t be necessary, it’s the best thing that could happen…

This adventure has not been in vain, we have met great people, and in any case we continue to develop the product, just in case!

2 – We continue !

Now that the prototype is well set up and – it seems – there is not so much urgency in France, although we are still waiting for the sources of the software by MIT, we will continue the development, it will still be useful for other countries.

In order to meet the specifications of a respirator as closely as possible, we searched for several hours on the sites of the Ministry of Health, of the National Public Health Agency, of the ARS (Regional Health Agencies), without finding any information on the Emergency Respirators !

It’s quite bizarre that no one in France has taken care of it while the Great Brittany and the USA have been working on this subject and that more than ten projects have started all over the world. The Bourgogne Franche-Comté ARS site indicates that they have 450 employees, if it’s everywhere like that in France (13 metropolitan regions), that’s a bit of a crowd anyway. And no one to deal with an essential subject such as Emergency Respirators?

The MHRA (Medicines and Healthcare products Regulatory Agency) in Great Britain has provided a complete and detailed document on Emergency Respirators minimum and recommended specifications, conditions of use, etc…
You can download the document by clicking here (in English)

This very well done document details the functions and safety features required for an unapproved Emergency Respirator and classifies them into 3 categories.

  • The Must Have : functions that the respirator must have
  • The functions that the vacuum cleaner should have, which the vacuum cleaner should have, are desirable, but not necessary, but not necessary.
  • Functions that the ventilator might have…

This Respirator complies with all the “Must Have” requirements requested by the MHRA, we have made some modifications to the product so that it meets the majority of the “Should Have” requirements in their document.

If it is not used in France, it will at least be possible to use it in Great Britain and in many other countries.

3 – The study is finished

Here are two computer-generated images of the final product.

We have designed the product so that it can be modulated according to the country of manufacture and according to the available possibilities.

Of course, in all versions, the 4 vital functions will be implemented :.

  • Respiratory rate (RF) breaths per minute: between 8 and 40.
  • Current volume (volume of air pushed into the lungs): between 200 and 800 ml depending on the weight and corpulence of the patient.
  • I/E ratio (ratio between inspiration and expiration time): it is recommended to start at about 1:2; it is possible to set it between 1:1 and 1:4.
  • Assisted detection pressure or trigger sensitivity : When a patient attempts to inhale, it can cause a drop in the order of 1 to 5 cm H2O, compared to the PEEP pressure (not necessary for the patient to breathe in).

He can also integrate :

  • Internal backup battery that gives it 2 hours of autonomy (for example during a transfer from hospital to hospital).
  • External battery socket – possibility to connect any car battery (charging or not on a car) or cigarette lighter, we think it could be useful in Africa.
  • Easy to clean and disinfect stainless steel cladding.
  • Ability to easily update the software and upgrade the ventilator.

But also:

  • Simplified version possible (without cowling using a very simple board to build with a standard Arduino Mega and the MIT files).

Once again:

This project is not a Respirator, but a Emergency ventilator, to be used by the hospital services in the emergency cases who will decide it and under their responsibility.

The device presented is not a medical device, it is not homologated, not certified by the Regional Health Agencies, not certified by the French Ministry of Solidarity and Health nor by a any organization.

It was carried out with the sole aim of compensating for an emergency situation where the equipment would be lacking and cannot replace certified equipment in terms of performance and reliability.

The sole purpose of this project is to help health care workers save lives.

4 – Our first prototype is currently running 24 hours a day.

It is about 100,000 breaths without any problem, except for a screw that has unscrewed (it will be necessary to mount it with the thread lock on the next assemblies). We are still far from the 700,000 breaths needed for a patient in CSR, but it is rather encouraging… The answer in about 12 days.

April 20th, 2020

Saturday 18, we received the parts of the 2nd prototype ventilator for the Valence team, they went to Lugny to recover the mechanics and the components.


  • This will help us a lot, 3 specialized engineers from Crouzet working on the project, it’s going to boost!
  • The tests on the ventilator continue, it is currently running 24 hours a day to detect any anomaly.
  • MIT has not yet provided the sources, but we are working on the engine control routines.
  • Hervé from Onwi has completed the ventilator map and launched the prototypes.


April 21st, 2020

  • Excellent news, we have just received confirmation from the Rotary Club of Mâcon, this organization will finance the pre-series of Emergency Respirators.

Logo Rotary Club Macon

  • Thank you to the Rotary club executive and board of directors for this generous contribution, which will accelerate the development of the product.
    Thanks to the energy of its executives, the decision was made in a matter of hours.
    This allowed us to immediately start the procurement and ordering of suppliers.
  • The project is supported by more and more companies. Hervé (again) found a great name for the project: “Respire“, the decision was not very democratic, I only had time to consult a few partners, but all those consulted agreed with this name. So it’s adopted: Respire becomes the official name of the project.
  • Hervé receives the printed circuit boards on Friday, and tests them right away! What energy!
  • Several large companies are expected to join the project, which should reduce the cost of pre-production runs. News as soon as possible.
  • Thanks to Eric for providing us with an Ambu Bag from another manufacturer. We have discovered a problem, there will be some adjustments to be made to the holders.

  • Otherwise, our prototype is still running… 24h/24

April 24th, 2020

  • Hervé received the pre-series of circuit boards…
    They are in the process of being assembled.
  • Automatic installation of SMD components

  • Soldering of SMD Components

  • Some pictures of the PCBa

Preserie RESPIRE

  • The prototype of the Valence team is running, it’s being tested.
    The measurements we’re making are different from those at MIT, so we’re wondering why we seem to need more power than MIT recommends.

  • The valence team will integrate the pressure sensor as quickly as possible to see if the Ambu is compressing too fast.

  • Our prototype is still running 24 hours a day at 100% air volume, with no visible wear and tear on the mechanics.

  • Our biggest concern : still no source of the software on the MIT website…

April 27th, 2020

  • Thanks to the help of the whole team, the first final version of “Respire ” has just been assembled tonight. We received the last pieces of ALPM and Hervé’s electronic card in the middle of the afternoon.
  • Tomorrow, we finish the wiring and the device should be running tomorrow evening. Kevin is coming to the offices, he will adapt the prototype software to this final configuration.

  • When you realize that less than a month ago, nothing existed! (apart from the work of MIT of course). What a job done by this incredible team of volunteers, what energy!

Respire final 003


    • Tomorrow, the Valence team is going up to Hervé in Lyon to validate the electrical consumption, the different functions of the electronic card, etc…
    •  I’m thinking of posting a video tomorrow of the product in operation as well as the Technical Specifications.

April 29th, 2020

Two excellent news today…

1 – MIT just released the sources for their respirator.

The Valence team is already adapting it, we hope to be able to start the tests on the 2nd or 3rd of May.

2-  The prototype is ready !

Today, the Valence team went to Hervé’s house in Lyon, where several control elements of the pcb were checked:

  • Air pressure sensor for air insufflated into the lungs
  • Speed and position control of the motor
  • Motor current consumption sensor
  • Validation push buttons
  • Temperature resistance tests of components
  • etc…

Once the software is adapted, the security tests will be able to start.

Mechanically, there is not much left to improve on the product, except the front panel markings and the security labels….

It should be possible to start the documentations and the different notices quickly.

It’s still a great coincidence, the day we’re ready, MIT will provide the code to control the respirator .

As soon as the sources have been adapted, we will be able to validate the operation of the potentiometers for adjusting the vital parameters:

  • Amount of air supplied to the patient,
  • Number of breaths per minute,
  • Limiting pressure of the air supplied to the patient,
  • Inspiration/expiration duration ratio.

Thanks to the whole team..

May 1st, 2020

Just some finishing touches while waiting for the adaptation of the MIT software.

A friend of ours made the decoration of the front panel of the ” Respire ” in order to better explain the functions.

Functions :


The aim is to be able to explain simply the functioning of the 4 vital functions of the ” Respire “, each vital function is adjusted by a potentiometer (a rotary knob) which adjusts the value displayed on the screen (one of the 2 upper lines of the screen). :

  • BPM (Breaths Per Minute or Respiration Rate) Breaths per Minute: The number of breathing cycles per minute (adjustable between 8 and 40 breaths per minute).
  • Tidal Volume: The volume of air breathed into the patient’s lungs. This volume varies according to the patient’s height and weight, usually between 200 and 800 ml of air.
  • I/E Ratio: The ratio of the patient’s Inhalation time to the patient’s Exhalation time. This ratio is usually set between 1/2 and 1/3, i.e. 1 inhalation time for 2 or 3 exhalation times, but can be adjusted from 1/1 to 1/4 on ” Respire “.
  • Trigger Volume: Helps to detect pressure or trigger sensitivity: for example when a patient tries to inhale, it can cause a fall of about 1 to 5 Cm/H2O, the device detects these anomalies.

More details on the MIT website:
Note: You may need to create an account on their site to access all information.

We have tried to simplify the operation as much as possible so that we can be intuitive in emergencies. The practitioner should not need any training to be able to immediately adjust the vital parameters of ” Respire “.

A graphic (yellow lines) indicates to the practitioner the display area of each vital function and the corresponding button.

http://www.spiderbot.eu/images/stories/Blog/covid19/face avant respire-01.jpg

When the practitioner wishes to change a value, simply set the value with the corresponding rotary knob and confirm with the “Validation” button. This prevents false operations and accidental changes of settings (vital in the case of a respirator).

Alarms :

MIT interviewed clinicians and reviewed existing ventilator standards to determine the major alarms needed, the appropriate actions to take when an alert condition is detected, and the behaviour expected to silence alarms.

The ” Respire ” will handle several alarm cases, the list of alarms is available here :   Détail des Alarmes 

In the event of a malfunction or a problem with the patient, the unit goes into alarm and triggers a beeper (more than 90 dB) plus a flashing red LED on the front panel.
An alarm can be silenced with the “Stop Alarm” button, this silence is valid for only 2 minutes.
The “Stop Alarm” button only silences the alarm tone, the visual notification remains displayed.
All alarms must be repaired within two minutes, even if a new alarm is triggered, the alarm silence must be maintained.
Pressing the “Stop Alarm” button again, before two minutes have elapsed, causes the alarm to stop.
If the same problem occurs again, the alarm will sound again.
If multiple alarms are active, they are visually displayed on the screen and the messages scroll.

This design of the front panel was dictated solely by the simplicity and functionality needed, if a Designer can help us make a nicer and more design product, he is welcome to!


May 6th, 2020

Integration of the MIT Code

After more than a week of work, we still don’t have a code that works!

In fact, it’s not that simple, MIT used a different motor driver than the one used at the beginning and this new driver simplifies their programming.

A “driver” is an electronic device which is able to drive a motor, there are different types of drivers according to their integration level, the one used by the MIT is among the most advanced, it integrates a microprocessor which manages the rotation, the directions, the encoder, the positions, the speeds, the accelerations?

Driver roboclaw

As a consequence, the MIT code doesn’t manage the engine at all, it just sends commands to the driver like: Move the arms to such position, at such speed and with such acceleration!  And the driver does all the calculations. 

The advantage of this solution is that it relieves the Arduino Mega of the most complex part of the calculations, it only manages the displays, the machine states and the alarms, and doesn’t take care of the engine at all…

Hervé had integrated the components to manage an engine but not such a complete and high level driver.

The Valence team worked for a week to integrate the engine management code, it is starting to work but there is still a lot of development work to be done.

We had a telephone meeting yesterday between the Valence team, Hervé and Lugny, we took the decision to use the same driver as MIT.

This leads to an additional cost of 80€ HT on the ventilator, but it will greatly simplify the development of the software, and above all, all the evolutions that MIT will bring in the future will be automatically integrated.

The use of this driver brings several advantages:

  • Possibility to use a large number of different motors
  • Possibility to use a large number of different encoders
  • Relatively simple change of motor type depending on availability (the market for DC motors seems to be a bit “tight” at the moment)

We ordered 2 drivers in the USA, we receive them on May 12th. So on May 13th everything should work.

Airway pressure measurement

In the meantime, we are working on the integration of the pressure sensor so that it can be integrated as close as possible to the mask or intubation tube.
The pressure sensor is essential for patient safety, it is used to measure the air pressure in the patient’s airway.
The airway pressure must be continuously monitored. The maximum pressure must be limited to 40 cm H2O. Other pressures are adjusted by the practitioner:

  • Positive End-Expiratory Pressure (PEEP) 5 – 20 cm H2O (will be realized by an external PEEP valve)
  • The plate pressure must be limited to a maximum of 30 cm H2O.

Due to the price of the pressure sensors and the support card, it is preferable that the pressure sensor is not disposable, after use, it must be located after a HEPA filter in order not to be contaminated by the patient’s air.


The pressure sensor will be inserted between the HEPA filter and the PEEP valve.

The aim is to obtain a robust component inserted in the ventilation line that is not likely to disconnect.

Of course, we will validate all this with people from reanimation unit as soon as we are ready.


12 thoughts on “Ventilator based on the MIT OpenSource project

    1. Hello,
      Thank you
      Unfortunately no, our software is not enough stable to be released.
      We made the prototype in less than 6 days and do not had enough time to develop the software.
      Our goal is to be ready “mechanically and electronically speaking” when the software of MIT will be released.
      An automatic ventilator needs a lot of security and alarms, not so easy to do, we are confident with the MIT team…
      We now have a team of about 10 people working on this project on a volunteer basis and we will provide all the designs, schema, wiring, BOM, Sources of software in OpenSource, but canno’t give any lead time today.
      We are well aware of the urgency, but can do no better at this time, sorry. Currently we can only provide the CAD files of the prototype in STEP (in French sorry, we have persons that currently works on it to translate in english).
      We have assembled the 1st prototype this afternoon, and discover some improvements to do, a fix will be on line as soon as possible.

      Take good care of your family and your health.


  1. Hi Guys. We have a similar device and are about to start writing some on the control code. What sensors do you have on the device? Pressure? Flow? Temp? Humidity? Please can you share the code even if it isnt stable. It would help a lot to have other people testing it. We could all share code.


    Wesley. RescueVent South Africa.

    1. Currently not, I am sorry.
      We do not have any sensor connected on it and the code does not handles the sensors, it is our next step…
      The prototype has been made with what we had available…
      Motor: CROUZET 8983b1, 12V CC, 15Nm, 30tr/mn
      Encoder is a Heindehain ROD 1020 3600 01 -03
      Looks like you’re further along than we are. Do you have any code that can help us?


  2. Guys! Thanks for an amazing work!
    I am building the same machine using your design.
    Could you please share the exact motor type you are using. Some links will help.

    1. We use a Crouzet 8983b1-2 12V reduction 45.
      Crouzet engineers are currently looking for a new motor that could be available easily with a higher reduction, because this one runs a bit too fast.
      It will change soon.
      You can download the latest STEP files of the “Respire” here :
      Same terms of use than the MIT, and by downloading, you agree to do not use these CAD files to built a emergency ventilator for humans.

  3. Thanks for you reply Philippe,
    what about Documentations for Hardware. ( electronics, screws, ball bearings…)
    Some sample codes etc.
    Is there a repository some where?

    1. It is still not finished, but I can send the prototype files by email, unfortunately your email does not works.
      Could you provide a valid email ?

  4. Dear Philippe,
    Could you suggest where we can get the motors online. Any alternative with compatible specs will also be useful for us.
    Thanks in advance

Leave a Reply

Your email address will not be published. Required fields are marked *