[HardDrive]

My wife is a physician, anesthesiologist who knows a bit about ventilation and airways.

I'm a software engineers, avid 3D printer, experienced with microcontrollers, blah, blah, blah.

We have a bit of time on our hands, so my wife and I really talked this thing through.

Mike is trying to drive the discussion towards introducing a device into our existing system, a product that would go through testing, approvals, etc. A welcome voice of reality, but missing the point. This is about the engineering exercise.

The original question: Could you build an open source ventilator.

The Harddrive engineering teams answer: For some limited functions of what a ventilator does, yes. But the machine could pose a very real threat to the patient if not calibrated correctly, and this gets back to the a VERY valid point: Without health care professionals trained on using the device, who understand its settings, it is unlikely to be of much good.

A ventilator is not a crude device that robotically pushes air in out. Simply pushing air in and out is not issue. A person can be 'breathing', but if the level of lung function has decreased to the point that air is not making it deep enough into the lungs to exchange oxygen, the mechanical action is irrelevant.

The magic in a ventilator is that you can dial in a specific pressure that will create enough force to reach the alveoli, yet not so much that you burst the person lungs. Could you 3D print a valve of sufficient quality to maintain a specific pressure? Yes, but 3D printing is done in plastic or resin. So even after you calibrate a valve, is it going to be accurate for 1000 cycles? 100,000? Unlikely.

Another strike against the homemade ventilator: Sensing a patient's breathing. If a person has simply stopped breathing, and the device can breath 100% for them, that is a simpler exercise. But what if the patient is still breathing, and you want to augment their natural breathing pattern. Could you build the sensors to detect a persons respiration. Yeah, sure, but let's get real here, are you going to leave a $10 arduino controller and a $2 crude mechanical sensor with someones life in their hands? No.

Conclusion: From a raw engineering perspective, can this be done: Yes. Can this be done in a way that anyone would actually stake the lives of their loved one on: Only in very, very desperate circumstances, and the potential for harm to the patient, or simply being ineffective is quite high without a person intimately familiar with both the medical problem and the engineering.