Ventilator Project Update: March 21th, 2020

Bruce Fenton

Decentralized Response

The Ventilator Project in conjunction with the New England Complex Systems Institute (NECSI) / is moving along and we wanted to provide an update to some of the recent activities of the group. There is still a lot to be done and we are just getting started, so if you or someone you know would be able to contribute, please get involved.

The goal is twofold: 1) to support existing manufacturers 2) to help makers with specs to built parts and fully functioning ventilators that can be used in their country or region.

As more and more volunteers are joining up, they are starting to break into smaller teams focused on specific activities/needs and we just wanted to post a few of those updates below as to what some individuals are working on, as well as additional sources of information regarding the project.

To learn more about Covid-19 and some helpful links be sure to visit the website at


If interested in helping out, please join our Slack channel via this link:

Public GitHub Repo on Existing Open Source Ventilator Projects — Robert L. Read:

We created a public GitHub repo ( to organize references to existing projects open source ventilator projects. Anyone can use a browser to see this or to request to have a project added by creating an issue.

Nariman Poushin and Robert L. Read used a meta-discussion channel to develop a process that focuses on testing and reliable to produce an informal certification of open source designs. This could be applied to many different designs. The goal is to allow medical professionals to have confidence in the clinical suitability, reliability, and trainability of the new designs to allow them to be deployed. Additionally, it may give donors enough confidence to support build teams to create hundreds, thousands, or tens of thousands of units. A diagram of this process and an explanation is in a public GitHub repo:

The demand modeling sub-team including Robert L. Read and James Trimarco began an effort to make an interactive ventilator demand/supply model to assess shortfalls.

Potential Deployment Roadmaps:

Suggestion #1:

Suggestion #2 by Anže Jarni:

The approach to developing a safe but affordable ventilator capable of keeping COVID-19 patients alive should be a result of organized teams led by team leads.

The keywords by priority are: safe, durable, cheep, user friendly

To develop such a product, several teams should be created based on their segment and role:

Medical team : A typical ventilator has many different modes. For the COVID-19 case the ventilator should be able to take care of pneumonia and even ARDS cases — this should be the current focus. In order for the SW and HW teams to understand what they need to build.

Software team should collaborate with the Medical team in order to implement the required algorithms.

Software Team should be implemented in modules, preferably in C language, because most of the microcontrollers speak the language. In case the software is written in C, complying with MISRA-C is a good idea, but should not be a requirement. Team leaders should define clear goals and tasks that follow specifications defined by the Medical team.

Hardware teams should collaborate with the Medical team in order to choose the right hardware in terms of quality and needs, such as continuous operation, portability for multiple weeks and sensitivity of sensors. The focus is to find parts that are appropriate but still widely available throughout the world.

Testing team should have medical knowledge and access to existing medical testing/learning tools, such as testing lungs. Testing team is responsible for ensuring that the final product works correctly and safely.

Certification team is assembled from a team knowledgeable in law and patents. Their responsibility is that the project does not become a financial burden, due to possible lawsuit of existing manufacturers and secondly to accelerate the standardization of the device

The project could be divided into the following phases:

Phase 1:

Medical team: prepares a document which SW and HW team can use

Software team: Starts UI development

Certification/patents team: does research on all existing patents

Phase 2:

Medical team: supports SW, HW team

Software team: Development of breathing algorithms and SW tests

Hardware team: Recognize and gather appropriate HW parts

Certification/patents team: collaborates to avoid patent case

Phase 3:

Medical team: tests the UI of the device and supports design of the device

Software team: Development of drivers for HW parts

Hardware team: Design of the device

Phase 4:

Medical team: extensive testing on artificial lung

Software team: final design touches, support to medical team

Hardware team: final design touches, support to medical team

Phase 5:

Certification team: Files for certification

Feedback Request for Physicians and Medical Device Manufacturers — From Isaac Sheets /
We are collecting feedback from Physicians and Medical Device Manufacturers. Please fill out the following form so we can receive your input and connect with you!

From Sarah McManus — An Exploration on ARDS:

I am writing a simple language exploration of ARDS, the very bad lung thing that can happen in severe cases of COVID-19. It has been reviewed & commented on by a molecular biologist and lung physiologist, Patricia Silveyra.

Google doc:

Why it would be useful to read this:

○ Understand the mechanisms of why “just give more oxygen” won’t work in severe cases, and why people are needing higher pressure full-on ICU ventilators — not Ambu bags or other low tech ventilation

○ Focus efforts on interventions that will be more effective (more ICU ventilators & training)

○ Share knowledge about the known causal factors in an easy-to-understand form

○ This will help people direct their energy and attention towards research on adjacent unknown areas that might unlock better treatments

Implications for ventilators

○ The manual bags (or low-tech mechanized versions) are not what people need when they have the specific kind of very bad lung thing that results from COVID-19.

○ Just moving air in & out at low pressure is not enough. Just oxygen (like through a nose tube) is not enough for severe cases. Apparently people need specific, finely adjustable, higher pressure ventilation in addition to oxygen.

○ This is because parts of the lung are getting damaged & collapsing on tiny and medium scales such that they cannot absorb oxygen or get rid of CO2 effectively.

From Alex Izvorski — Available Specifications for Open Source Ventilators:

UK Rapidly Manufactured Ventilation System(RMVS) Spec:

Canada Spec:

Unofficial spec:

Julian Botta — Specifications for simple open source mechanical ventilator

The response-ventilator team has also been looking into making available simple gas powered (operating off of the pressure from an oxygen cylinder or supply) ventilators. They are recognized as an easy to manufacture low cost disposable emergency alternative to complex hospital units. They are potentially 3d printable (plus some hardware, diaphragms and tubing) and relatively simple to assemble. E.g. Vortran Go2Vent and similar. There are reference designs from expired patents and a literature background of existing products with testing data. Outreach to existing manufacturers to release designs is a possibility. More information about this can be found at

Overall, the team of volunteers is still coming together and organizing itself into smaller working groups focused on specific tasks and the best way to work towards common goals. As we continue to work to build this out, any additional help is greatly appreciated and welcomed. If you know of someone who could contribute to this, please feel free to pass along this article and all relevant links to them.

To learn more about Covid-19 and some helpful links be sure to visit the website at

If interested in helping out, please join our Slack channel via this link:

I will continue to post updates below and use this post as a central location to dispense information.