The State of Open Source Ventilator Projects as of March 21st

Robert L. Read
4 min readMar 23, 2020

— Robert L. Read, PhD, Nariman Poushin, Keeshan Patel, Stacey Oliver Sarris and Andrea Ippolito

Diagram of various test steps for Open Source Ventilators

At this moment…

There are thousands of intelligent, diligent, well-meaning engineers trying to help the design of open source ventilators to address a possibly imminent life-threatening shortage caused by the COVID-19 pandemic. This wealth of creative technical energy is currently disorganized, scattered, and unfocused. Rather than being a tremendous force for saving lives that it may become, energy and time is currently being wasted on oversimplifications of the problem and the belief that the projects are closer to deployable than they really are.

We have been working on this full-time only for the last eight days. With great humility, we would like to assert a few things.

There are Few Open Projects Ready for Volunteers

Right now there are four projects better defined than others:

The situation may change quickly, but right now these four projects are the only projects that are even minimally actionable.

Other closed projects have been reported in the news and we wish them well, but if a project has not published a design, it cannot be considered a real open source project welcoming volunteers.

At (NECSI) we have compiled a list of projects, both open and closed, and resources which we invite you to correct and amend by entering an issue.

The Complexity of the Problem is Underestimated

At least some engineers seem to have undertaken or thought about undertaking projects without sufficient understanding of the problem. We consider the four links below essential reading. They are about 50 pages in all, and should take at most two hours to read:

People are rightly focused on prototypes, but must soon focus on testing. Without extensive, rigorous testing, no ventilator can be deployed.

Testing will be 90% of the effort in saving lives through open source ventilators.

Eight days ago we did not know what ventilator induced barotrauma is. Now we know: if a ventilator puts a little too much pressure on a sick person’s lungs, they will be injured in their weakest hour and may die. The same is true of too much volume. You cannot easily blow too much air into a strong, conscious person’s lungs, but you can easily kill a weak, unconscious person on a ventilator by doing so. Patients in general need many forms of breathing assistance, such as simple supplemental oxygen. But COVID-19 patients sometimes have ARDS and require complete intensive ventilation for up to a week or more in an ICU. A machine to do that safely is MUCH harder to make than a machine that moves air periodically. This is why our team at (NECSI) have specifically emphasized testing, and proposed a systematic testing process that might give a clinician sufficient confidence to deploy an open source ventilator.

Weighing The Risk

That extreme risk and life-critical nature makes DIY, open source ventilator designs a solution of last resort. Let us emphasize that: LAST RESORT. When an open source ventilator is needed, if it is not fully tested, reliable, and easy-to-use by a clinician wearing personal protective equipment, we risk costing lives rather than saving them.

How to Help

But the last resort may come upon us, so we are supporting the design, testing, and manufacture of open source ventilators. We have slowly learned how difficult this task will be, and what a challenge it will be to do quickly. But it is worth doing, and it may need your help.

If you are a firm that can manufacture electromechanical devices:

  1. Please see the University of Cambridge’s recommendations.

If you are an engineer:

  1. Do your homework.
  2. Offer to help an existing project.
  3. Focus on testing and reliability.

If you feel called to design your own ventilator (and we do need more designs):

  1. Do your homework.
  2. If you build something, publish the design fully with open licenses, even as you are developing it. If there is no published design, it is useless to the community.
  3. Plan to have others build your design as a test of your design documentation.
  4. Think in terms of manufacturing units on at the scale of hundreds or thousands or more. Assume that others will assist you financially in this if you publish a clear, well-tested design.

If you are a donor:

  1. Thank you! Buy equipment and supplies for teams that are open and have done their homework.

If you are a medical professional:

  1. Be patient with the engineers, but educate them on risks and critical performance features. Join a project and provide guidance.
  2. Help ALL of the community understand how to seriously address this crisis.



Robert L. Read

Public Inventor. Founder of Public Invention. Co-founder of @18F. Presidential Innovation Fellow. Agilist. PhD Comp. Sci. Amateur mathematician.