Whole world is waiting for blueprints. There are many many industrial 3D printers out there. Even with fraction of them companies could print all the needed ventilators worldwide. There is no worldwide authority, that can provide these blueprints. And that’s imho the biggest problem today.
Edit: It might contain electronics, but there is pandemic, not global war, so printed circuit manufacturing capacity is still there. Some modules with microprocessor are widely available worldwide and the design can be constrained using such module and 2 layer carrier printed circuit board. Machines to build these are available at better maker spaces. Pick and place machines are also there. So basically my timeline would as follows: order parts on Digikey and mouser today, go to maker space, start engraving printed circuit boards and launch 3D printers. On Monday or Tuesday digikey ships the parts. Maybe it will arrive on Wednesday, shipping companies are busy right now. On Wednesday I would launch pick and place machine and produce first batch of boards and next day I would start assembling manually the devices. On next weekend I would have probably 10 units to start tests. That’s working alone. With other tech savvy individuals output would be orders of magnitude higher. And this is not theory, in my previous workplace team of 8 blue collar workers was shipping 2000 units of comparable machinery per year.
Every ventilator manufactured can provide blue prints of their old designs. There has to be old design without microcontrollers that could be made in most machine workshops around the world. I guess a modern design would be to complicated to outsource to random machine shops with all the special electronics.
I've been following the developments of these ventilators pretty closely, particularly of these using a ventilator for multiple patients.
The hard part is not the ventilation mechanism itself, but the specifications.
Those ventilators must be controlled precisely, they must take into account the characteristics of COVID-19 patients. There is some overlap with ventilation for ARDS. The more you read into this subject, the more you realize how absolutely scary it is to actually design such a project.
Lots of parameters have to be taken into account.
Yeah, I' been following the various projects, too.
As far as I can see, many people underestimate that ventilators are just one necessary tool of many in intensive care.
Sure, you can whip up a proof-of-concept prototype using a bag-valve mask, a stepper motor, and an Arduino.
However, most COVID-19 patients will be multimorbid and require lots of attention.
As multiple healthcare experts have mentioned so far, the real bottleneck are highly-qualified intensive care doctors and nurses.
Do you really want to give these people a new ventilator with which they are not familiar, which may lack features they need, and which might break down or malfunction in an unfortunate moment?
Maybe there is a place for such simple ventilators outside intensive care, e.g., as support for people who are able to breathe on their own and do not necessarily require mechanical ventilation, AND who are in a place without nasal O2 supply?
If someone didn't need mechanical ventilation but would benefit from some assistance, couldn't they use a CPAP? There sure are a lot of those in use across the world.
Boeing got FAA approval for aviation-grade 3D parts during the 787 program.
I don't know if they make them in-house, through a supplier, or both, but they have the capacity to not only get the things manufactured but deliver them too.
It is not a ventilator but the blower motor and various sensors are easily available on Aliexpress but it is the software that decides the correct airflow pressure at the right time when an apnea event occurs that is the secret sauce that needs to be clinically approved.
I presume the same for a ventilator - the software that helps the respiratory therapist to monitor the patient.
There was a conversation recently on HN that I can’t find just now, where an expert claimed a DIY ventilator project will most likely only create a ‘glorified CPAP machine’ and that real hospital ventilators are much more complicated and require specially trained people to use properly, as they can be very dangerous.
Not to dissuade anyone from trying to make open source ones a reality in the long run though...
CPAP machines are probably still needed more in this pandemic, there will be a lack of them as well, and having more available will save lives. (I'm not a doctor.) AFAIK they've run out of even basic medical oxygen delivery capacity in many places in Italy, which is even simpler than CPAP.
ResMed is a big manufacturer of CPAP machines as well as ventilators, and is the only US-based company listed in this Reuters article about ramping up ventilator production [0]. It says, "ResMed has reallocated resources away from producing other devices to meet demand that is more than four times normal levels". So it's possible they are making fewer CPAP machines and more ventilators.
The electronics and software would be a critical component as well. It really wouldn't cost that much to develop the software/hardware or even the electronics as microchips are now often cheaper than an equivalent set of discrete analog components. The _real_ costs in any medical equipment is passing FDA (or equivalent European) certifications.
The most plausible way to handle these situations would be to have "pre-authorized" set of plans, bill of materials (or qualified components), and manufacturing process. Critically there would need to some sort of checklist and verification process to verify the equipment (preferably able to be performed by a third-party with minimal training). A consortium of FDA, WHO, etc worked with a established non-profit organization could design a set of plans and procedures for "emergency medical equipment". Then the entire world could be prepared for such emergencies, and possibly double as lower cost equipment for developing nations.
Likely established manufacturers would lobby hard against such measures (why would hospitals pay $50-60k for a ventilator when these "open plan" units only cost $1k?). There would also be questions of reliability of such devices. These issues could be mitigated by authorizing use of such equipment only during state of emergencies. This would make sense as such equipment would possibly be of lower (or just uncertain) quality compared to normal devices. But the laws allowing for this would need to be passed, and again, that's the difficult part. The technology/manufacturing/3d printing is almost inconsequential _without_ appropriate legal infrastructure in place.
I absolutely share your opinion. Business of the medical equipment companies shouldn’t canibalized. But they can’t deliver equipment when needed, so alternative solution must be found. There also probably countries, that cannot afford $50-60k ventilators anyway. So for these countries or all countries during pandemic easy manufacturable equipment should be designed.
Everyone likes 3D printing, but other than giving them some more love, is there a reason that they wouldn't use real CNC machines that could make parts much, much faster? I'd guess that VW probably has a few.
I've been wondering this myself. Every nerd, hacker, and maker space are talking about 3d printing our way out of this mess, but I've used a 3d printer before. They're not quick.
It's pretty easy to build a diy injection molder. Why aren't the hobbiests with cnc lathes and mills making molds and sending them to the hobbiests without?
For many geometries (especially complex or thin-walled shapes), 3D printing is faster than milling.
Many medical devices are made largely out of plastic, especially the disposable parts that are currently in short supply. 3D printers (FDM and SLS) commonly produce parts in many of these same plastics. Mills are generally used on metal, and the programmers and operators may not be as comfortable fabricating complex shapes from plastic. Mills are especially slow making thin plastic pieces, as they have to limit the feed- and pull-out-forces, and the heat generated, by the tool.
3D printers have little difficulty in making flexible / compliant parts (like close-fitting masks), because they exert nearly zero force on parts as they make them. Mills are even slower at making flexible plastic parts, because they have to operate at even higher RPMs, lower feed rates, and shallower depth-of-cuts in order to not deform the flexible part, even as they rip material off of it.
3D printers convert nearly all of the material you give them in to finished parts. Mills produce a lot (sometimes an awful lot) of waste that has to be shipped away and recycled back into usable material.
VW probably relies on CNC mills in their normal production pipeline, but may not have hard dependencies on their 3D printers. It might be much less disruptive for them to offer much or all of their 3D printing capacity to humanitarian efforts.
---
The highest-output way to utilize CNC mills for bridging the shortage of ventilator / PPE parts would be to fabricate molds for injection-molded parts. VW (and other automakers) can definitely do this. The big holdups would be in acquiring designs for the parts (who knows how long that would take?), and in designing the molds (on the order of a week or two). Milling a mold takes hours or days depending on how complex it is. (The different pieces of a mold can be made by different machines.) Once fabricated, assembled, and inspected, a mold could produce hundreds of parts a day.
The necessary delay in designing, fabricating, and iterating molds (and other assemblies for the mass-manufacture of plastic parts) means that 3D printing will always beat it to the punch. For the next couple of weeks, expect most unofficially-manufactured medical parts to be 3D printed.
The supply of traditionally-manufactured medical parts should significantly increase later this year. The parts and assemblies to do that are/will-be made on CNC mills in the intermediate time.
Thanks but looking at the various ISO connector specs and the fact that oxygen at pressure is involved , don't think a hacker project using a raspberry pi running Raspbian is going to cut it.
Just to point out, the members of some hackspaces around the world are a bit more "proper engineering" oriented + experienced with real world industrial design than you're alluding to.
Not everyone is still into low grade crap / toys. Although they're a useful starting point when first getting into making things. ;)
What would be interesting to understand is how do they bootstrap the production process and what design methodology they use for the ventilator.
Do they just go to Siemens and ask for a basic schematics they are happy to share that could potentially pass the certification process?
How does the factory organises itself to change the production lines. I'm fascinated about this topic.
I'm a huge 3D printer advocate and hobbyist but...this doesn't really make sense to me. VW (and arguably a lot German based companies) are super in-sourced based. I wouldn't be surprised they have a dedicated plant just for injection molding and they probably make their own tooling in-house. I don't see how 3D printing would be the best choice for their infrastructure. They probably could get usable test shots within a week with the right people working around the clock.
37 comments
[ 2.7 ms ] story [ 83.7 ms ] threadEdit: It might contain electronics, but there is pandemic, not global war, so printed circuit manufacturing capacity is still there. Some modules with microprocessor are widely available worldwide and the design can be constrained using such module and 2 layer carrier printed circuit board. Machines to build these are available at better maker spaces. Pick and place machines are also there. So basically my timeline would as follows: order parts on Digikey and mouser today, go to maker space, start engraving printed circuit boards and launch 3D printers. On Monday or Tuesday digikey ships the parts. Maybe it will arrive on Wednesday, shipping companies are busy right now. On Wednesday I would launch pick and place machine and produce first batch of boards and next day I would start assembling manually the devices. On next weekend I would have probably 10 units to start tests. That’s working alone. With other tech savvy individuals output would be orders of magnitude higher. And this is not theory, in my previous workplace team of 8 blue collar workers was shipping 2000 units of comparable machinery per year.
Every ventilator manufactured can provide blue prints of their old designs. There has to be old design without microcontrollers that could be made in most machine workshops around the world. I guess a modern design would be to complicated to outsource to random machine shops with all the special electronics.
Nor should there be. Centralization isn't beneficial for this sort of problem.
Or the hardest part are the 3D printable parts?
The hard part is not the ventilation mechanism itself, but the specifications. Those ventilators must be controlled precisely, they must take into account the characteristics of COVID-19 patients. There is some overlap with ventilation for ARDS. The more you read into this subject, the more you realize how absolutely scary it is to actually design such a project. Lots of parameters have to be taken into account.
Read on mechanical ventilation. https://en.wikipedia.org/wiki/Mechanical_ventilation
3D printed parts that where the air flow through must be air-tight.
Do you really want to give these people a new ventilator with which they are not familiar, which may lack features they need, and which might break down or malfunction in an unfortunate moment?
Maybe there is a place for such simple ventilators outside intensive care, e.g., as support for people who are able to breathe on their own and do not necessarily require mechanical ventilation, AND who are in a place without nasal O2 supply?
I don't know if they make them in-house, through a supplier, or both, but they have the capacity to not only get the things manufactured but deliver them too.
It is not a ventilator but the blower motor and various sensors are easily available on Aliexpress but it is the software that decides the correct airflow pressure at the right time when an apnea event occurs that is the secret sauce that needs to be clinically approved.
I presume the same for a ventilator - the software that helps the respiratory therapist to monitor the patient.
Not to dissuade anyone from trying to make open source ones a reality in the long run though...
This is all quite true, however from the sound of things places like New York City need both ventilators and staff.
[0] https://www.reuters.com/article/us-health-coronavirus-ventil...
The most plausible way to handle these situations would be to have "pre-authorized" set of plans, bill of materials (or qualified components), and manufacturing process. Critically there would need to some sort of checklist and verification process to verify the equipment (preferably able to be performed by a third-party with minimal training). A consortium of FDA, WHO, etc worked with a established non-profit organization could design a set of plans and procedures for "emergency medical equipment". Then the entire world could be prepared for such emergencies, and possibly double as lower cost equipment for developing nations.
Likely established manufacturers would lobby hard against such measures (why would hospitals pay $50-60k for a ventilator when these "open plan" units only cost $1k?). There would also be questions of reliability of such devices. These issues could be mitigated by authorizing use of such equipment only during state of emergencies. This would make sense as such equipment would possibly be of lower (or just uncertain) quality compared to normal devices. But the laws allowing for this would need to be passed, and again, that's the difficult part. The technology/manufacturing/3d printing is almost inconsequential _without_ appropriate legal infrastructure in place.
It's pretty easy to build a diy injection molder. Why aren't the hobbiests with cnc lathes and mills making molds and sending them to the hobbiests without?
Many medical devices are made largely out of plastic, especially the disposable parts that are currently in short supply. 3D printers (FDM and SLS) commonly produce parts in many of these same plastics. Mills are generally used on metal, and the programmers and operators may not be as comfortable fabricating complex shapes from plastic. Mills are especially slow making thin plastic pieces, as they have to limit the feed- and pull-out-forces, and the heat generated, by the tool.
3D printers have little difficulty in making flexible / compliant parts (like close-fitting masks), because they exert nearly zero force on parts as they make them. Mills are even slower at making flexible plastic parts, because they have to operate at even higher RPMs, lower feed rates, and shallower depth-of-cuts in order to not deform the flexible part, even as they rip material off of it.
3D printers convert nearly all of the material you give them in to finished parts. Mills produce a lot (sometimes an awful lot) of waste that has to be shipped away and recycled back into usable material.
VW probably relies on CNC mills in their normal production pipeline, but may not have hard dependencies on their 3D printers. It might be much less disruptive for them to offer much or all of their 3D printing capacity to humanitarian efforts.
---
The highest-output way to utilize CNC mills for bridging the shortage of ventilator / PPE parts would be to fabricate molds for injection-molded parts. VW (and other automakers) can definitely do this. The big holdups would be in acquiring designs for the parts (who knows how long that would take?), and in designing the molds (on the order of a week or two). Milling a mold takes hours or days depending on how complex it is. (The different pieces of a mold can be made by different machines.) Once fabricated, assembled, and inspected, a mold could produce hundreds of parts a day.
The necessary delay in designing, fabricating, and iterating molds (and other assemblies for the mass-manufacture of plastic parts) means that 3D printing will always beat it to the punch. For the next couple of weeks, expect most unofficially-manufactured medical parts to be 3D printed.
The supply of traditionally-manufactured medical parts should significantly increase later this year. The parts and assemblies to do that are/will-be made on CNC mills in the intermediate time.
I hope I am wrong but it seems rather daunting.
Not everyone is still into low grade crap / toys. Although they're a useful starting point when first getting into making things. ;)
Do they just go to Siemens and ask for a basic schematics they are happy to share that could potentially pass the certification process? How does the factory organises itself to change the production lines. I'm fascinated about this topic.
There is an open source effort to create a ventilator that can be produced in high volume: https://www.notion.so/1-Million-Ventilators-1035c5cf946445a7...
What is VW's approach to this and how do they plan for it. Grateful if anyone has more insight into this topic.
An "iron lung" seems like something a lot easier to design in such a way that it won't harm people.