eevBLAB #72 - BUSTED! Medtronics Open Ventilator FAIL

[EEVblog]

Unfortunately Medtronics opening up their design files for their Puritan Bennett 560 Ventilator is not much more than a publicity stunt.

[navin_siri]

Hey Dave, I think this is somewhat unfair. The word "Open Source" is not used on the company website, which I think would give a clear insight to the intentions of the company to not intending to actually be open source. If they actually wanted to be open source, wouldn't they plaster that word everywhere?

[johnlsenchak]

This is just speculation  on my part  , but  maybe  a company  like Medtronics  is being careful  not  putting  all the technical  information   online  due to the fact that a rogue  country  like China  would use them  for profit  The country  is well known  known  for stealing  documents  for the purpose  of  creating  counterfeit  products  of  low quality


If  push came to shove  and you  where in need of  ventilator   would  you trust   one  from  made in  China ? I wouldn't

[EEVblog]

Hey Dave, I think this is somewhat unfair. The word "Open Source" is not used on the company website, which I think would give a clear insight to the intentions of the company to not intending to actually be open source. If they actually wanted to be open source, wouldn't they plaster that word everywhere?

They have "open" in the URL Medtronic.com/openventilator
In their press release:

“By openly sharing the PB 560 design information, we hope to increase global production of ventilator solutions for the fight against COVID-19.”

Sorry, I'm calling BS on that.
Sure they aren't saying "open source", but that's not the point, they are clearly trying to give the impression that they are saving the world by releasing "design information", in full knowledge that the media are going to run with this as it seems they have done.
What they have instead released of any real value is the requirements documents, and only then really useful to a multi-year approved product design effort (I stand to be corrected on that). It would do very little or nothing to address any need for ventilators now.
And even then you've gotta read the permissive license very carefully, as the license ends the final day the WHO’s Pandemic call is in effect is in effect, or October 1, 2024, whichever comes first.

[Brumby]

Sure they aren't saying "open source", but that's not the point, they are clearly trying to give the impression that they are saving the world by releasing "design information", in full knowledge that the media are going to run with this as it seems they have done.

Which is what you would expect from the Media.

Details, accuracy, practicalities and other engineering aspects are totally irrelevant.  Their job is to attract people's attention - so they can sell their audience to advertisers.  Actually saving lives would be a bonus story - but they won't be betting on that one.  It's enough that they can sell the idea of that happening.

[Mr.B]

A 100% Medtronics publicity stunt.
Deliberately manipulating the dumb media to pull on the heartstrings of the common public to appear that they alone will save the world.
I hope their share value dives and the fleas of a thousand camels infest their patents.

[EEVblog]

Now, the cynic in me might also think it may have something to do with how Medtronic failed to deliver ventilator for a US government contract since 2015:
https://www.nytimes.com/2020/03/29/business/coronavirus-us-ventilator-shortage.html

Thirteen years ago, a group of U.S. public health officials came up with a plan to address what they regarded as one of the medical system’s crucial vulnerabilities: a shortage of ventilators.

The breathing-assistance machines tended to be bulky, expensive and limited in number. The plan was to build a large fleet of inexpensive portable devices to deploy in a flu pandemic or another crisis.

Money was budgeted. A federal contract was signed. Work got underway.

And then things suddenly veered off course. A multibillion-dollar maker of medical devices bought the small California company that had been hired to design the new machines. The project ultimately produced zero ventilators.

That failure delayed the development of an affordable ventilator by at least half a decade, depriving hospitals, states and the federal government of the ability to stock up. The federal government started over with another company in 2014, whose ventilator was approved only last year and whose products have not yet been delivered.

Today, with the coronavirus ravaging America’s health care system, the nation’s emergency-response stockpile is still waiting on its first shipment. The scarcity of ventilators has become an emergency, forcing doctors to make life-or-death decisions about who gets to breathe and who does not.

The stalled efforts to create a new class of cheap, easy-to-use ventilators highlight the perils of outsourcing projects with critical public-health implications to private companies; their focus on maximizing profits is not always consistent with the government’s goal of preparing for a future crisis.


ImageThomas R. Frieden, the former director of the Centers for Disease Control and Prevention. “We definitely saw the problem,” he said.
Thomas R. Frieden, the former director of the Centers for Disease Control and Prevention. “We definitely saw the problem,” he said.Credit...Chip Somodevilla/Getty Images
“We definitely saw the problem,” said Dr. Thomas R. Frieden, who ran the Centers for Disease Control and Prevention from 2009 to 2017. “We innovated to try and get a solution. We made really good progress, but it doesn’t appear to have resulted in the volume that we needed.”

The project — code-named Aura — came in the wake of a parade of near-miss pandemics: SARS, MERS, bird flu and swine flu.

Federal officials decided to re-evaluate their strategy for the next public health emergency. They considered vaccines, antiviral drugs, protective gear and ventilators, the last line of defense for patients suffering respiratory failure. The federal government’s Strategic National Stockpile had full-service ventilators in its warehouses, but not in the quantities that would be needed to combat a major pandemic.

Latest Updates: Markets and Business
Asian markets rise modestly despite Wall Street’s surge.
Wall Street climbs but investor unease continues.
New York’s attorney general has questions for Zoom about privacy and security.
See more updates
Updated 43m ago
More live coverage: Global New York
In 2006, the Department of Health and Human Services established a new division, the Biomedical Advanced Research and Development Authority, with a mandate to prepare medical responses to chemical, biological and nuclear attacks, as well as infectious diseases.

In its first year in operation, the research agency considered how to expand the number of ventilators. It estimated that an additional 70,000 machines would be required in a moderate influenza pandemic.

The ventilators in the national stockpile were not ideal. In addition to being big and expensive, they required a lot of training to use. The research agency convened a panel of experts in November 2007 to devise a set of requirements for a new generation of mobile, easy-to-use ventilators.

In 2008, the government requested proposals from companies that were interested in designing and building the ventilators.

The goal was for the machines to be approved by regulators for mass development by 2010 or 2011, according to budget documents that the Department of Health and Human Services submitted to Congress in 2008. After that, the government would buy as many as 40,000 new ventilators and add them to the national stockpile.


Image
The federal government has spent years trying to increase the supply of ventilators in its strategic stockpile.
The federal government has spent years trying to increase the supply of ventilators in its strategic stockpile.Credit...Cooper Neill for The New York Times
The ventilators were to cost less than $3,000 each. The lower the price, the more machines the government would be able to buy.

Companies submitted bids for the Project Aura job. The research agency opted not to go with a large, established device maker. Instead it chose Newport Medical Instruments, a small outfit in Costa Mesa, Calif.

Newport, which was owned by a Japanese medical device company, only made ventilators. Being a small, nimble company, Newport executives said, would help it efficiently fulfill the government’s needs.

Sign up to receive our daily Coronavirus Briefing, an informed guide with the latest developments and expert advice.
Sign Up
Ventilators at the time typically went for about $10,000 each, and getting the price down to $3,000 would be tough. But Newport’s executives bet they would be able to make up for any losses by selling the ventilators around the world.

“It would be very prestigious to be recognized as a supplier to the federal government,” said Richard Crawford, who was Newport’s head of research and development at the time. “We thought the international market would be strong, and there is where Newport would have a good profit on the product.”

Federal officials were pleased. In addition to replenishing the national stockpile, “we also thought they’d be so attractive that the commercial market would want to buy them, too,” said Nicole Lurie, who was then the assistant secretary for preparedness and response inside the Department of Health and Human Services. With luck, the new generation of ventilators would become ubiquitous, helping hospitals nationwide better prepare for a crisis.

The contract was officially awarded a few months after the H1N1 outbreak, which the C.D.C. estimated infected 60 million and killed 12,000 in the United States, began to taper off in 2010. The contract called for Newport to receive $6.1 million upfront, with the expectation that the government would pay millions more as it bought thousands of machines to fortify the stockpile.

Project Aura was Newport’s first job for the federal government. Things moved quickly and smoothly, employees and federal officials said in interviews.

Every three months, officials with the biomedical research agency would visit Newport’s headquarters. Mr. Crawford submitted monthly reports detailing the company’s spending and progress.

The federal officials “would check everything,” he said. “If we said we were buying equipment, they would want to know what it was used for. There were scheduled visits, scheduled requirements and deliverables each month.”

In 2011, Newport shipped three working prototypes from the company’s California plant to Washington for federal officials to review.

Dr. Frieden, who ran the C.D.C. at the time, got a demonstration in a small conference room attached to his office. “I got all excited,” he said. “It was a multiyear effort that had resulted in something that was going to be really useful.”

In April 2012, a senior Health and Human Services official testified before Congress that the program was “on schedule to file for market approval in September 2013.” After that, the machines would go into production.

Then everything changed.

The medical device industry was undergoing rapid consolidation, with one company after another merging with or acquiring other makers. Manufacturers wanted to pitch themselves as one-stop shops for hospitals, which were getting bigger, and that meant offering a broader suite of products. In May 2012, Covidien, a large medical device manufacturer, agreed to buy Newport for just over $100 million.

Covidien — a publicly traded company with sales of $12 billion that year — already sold traditional ventilators, but that was only a small part of its multifaceted businesses. In 2012 alone, Covidien bought five other medical device companies, in addition to Newport.

Newport executives and government officials working on the ventilator contract said they immediately noticed a change when Covidien took over. Developing inexpensive portable ventilators no longer seemed like a top priority.

Newport applied in June 2012 for clearance from the Food and Drug Administration to market the device, but two former federal officials said Covidien had demanded additional funding and a higher sales price for the ventilators. The government gave the company an additional $1.4 million, a drop in the bucket for a company Covidien’s size.

Government officials and executives at rival ventilator companies said they suspected that Covidien had acquired Newport to prevent it from building a cheaper product that would undermine Covidien’s profits from its existing ventilator business.

Some Newport executives who worked on the project were reassigned to other roles. Others decided to leave the company.

“Up until the time the company sold, I was really happy and excited about the project,” said Hong-Lin Du, Newport’s president at the time of its sale. “Then I was assigned to a different job.”


Image
Hong-Lin Du was the president of Newport Medical Instruments, a small company in Costa Mesa, Calif., at the time of its sale to Covidien.
Hong-Lin Du was the president of Newport Medical Instruments, a small company in Costa Mesa, Calif., at the time of its sale to Covidien.Credit...Tag Christof for The New York Times
In 2014, with no ventilators having been delivered to the government, Covidien executives told officials at the biomedical research agency that they wanted to get out of the contract, according to three former federal officials. The executives complained that it was not sufficiently profitable for the company.

The government agreed to cancel the contract. The world was focused at the time on the Ebola outbreak in West Africa. The research agency started over, awarding a new contract for $13.8 million to the giant Dutch company Philips. In 2015, Covidien was sold for $50 billion to another huge medical device company, Medtronic. Charles J. Dockendorff, Covidien’s former chief financial officer, said he did not know why the contract had fallen apart. “I am not aware of that issue,” he said in a text message.

Robert J. White, president of the minimally invasive therapies group at Medtronic who worked at Covidien during the Newport acquisition, initially said he had no recollection of the Project Aura contract. A Medtronic spokeswoman later said that Mr. White was under the impression that the contract had been winding down before Covidien bought Newport.

In a statement Sunday night, after the article was published, Medtronic said, “The prototype ventilator, developed by Newport Medical, would not have been able to meet the specifications required by the government, nor at the price required.” Medtronic said that one problem was that the machine was not going to be usable with newborns.

It wasn’t until last July that the F.D.A. signed off on the new Philips ventilator, the Trilogy Evo. The government ordered 10,000 units in December, setting a delivery date in mid-2020.

As the extent of the spread of the new coronavirus in the United States became clear, Dr. Anthony S. Fauci, director of the National Institute of Allergy and Infectious Diseases, revealed on March 15 that the stockpile had 12,700 ventilators ready to deploy. The government has since sped up maintenance to increase the number available to 16,660 — still fewer than a quarter of what officials years earlier had estimated would be required in a moderate flu pandemic.

Last week, the Health and Human Services Department contacted ventilator makers to see how soon they could produce thousands of machines. And it began pressing Philips to speed up its planned shipments.

The stockpile is “still awaiting delivery of the Trilogy Evo,” a Health and Human Services spokeswoman said. “We do not currently have any in inventory, though we are expecting them soon.”

[Bud]

Do not know why they call it open source. To me the message was clear when i saw :

"Our specification, your ingenuity"

Their specification. Specification. Get it?

[Smokey]

I haven't looked at these documents to see what's exactly in them, but it sounds like there is enough to essentially make/use their specification for this machine.

Having an established proven working specification is a HUGE time saver for an engineering effort.  Anyone that wants to make themselves a device now would have the roadmap.  The spec if the map.  If you are actually capable of doing the engineering work across multiple disciplines or leading a team, which most people aren't, then this would be enough to get you going.

I don't design ventilators, so if you asked me to make one from scratch I would need to first make a detailed specification to ensure I really understood all the important aspects of the problem at hand.  As any engineer knows that has ever had to go through this process, you cut corners on the spec at your own peril.  When you have that "You mean I need to monitor the oxygen concentration of the exhaled air?  I didn't think of that! No one told me that!  I need to do another rev!" moment, you are going to wish you had a proven design spec to work from at the beginning.

[Smokey]

::Rant on::
I feel like this is another example of why "open hardware", especially whole complicated product designs, is to a large extent a lie anyway. 
(By lie I mean having a chance at recreating the device by yourself)

Take just one piece here.  Say they gave you the cad files for the injection molds, AND you happen to either be an injection mold expert or have an amazingly excellent relationship with an injection molding company that is willing to stop all their other paying work to help you try to make a ventilator.  Great.  How long does it take to have the dies machined for your specific injection molding machine?  How long to get the dies modified after they come back and don't work the first time (and likely second time)?  How much machine time and material cost can you dedicate to this side project?  And this is all assuming you already knew what you were doing.  Your average garage hacker has a zero chance of success here. 

The same thing goes for essentially every part of a machine like this.  Metal parts need a machine shop, which means stuff like fixtures and CNC programs (more iterations).  Electrical parts need PCBs and components sourced, and stencils, and pick and place programs, and bed of nails fixtures (more iterations).  You probably need a fairly complicated final functional test setup that involves references like specific O2 concentrations and calibrated flow rates (more iterations).  All these manufacturing processes are iterated over many revs before the bugs are worked out at a serious cost of time and money and man hours.  What's the MTBF?  Once a design is rolling in production you don't make any non-essential changes (part substitutions, vendor switches, skip procedures) because you can't foresee all the consequences but you think some garage hacker can get all this working in a short period of time.  This thing needs to run 24/7 and CAN NOT FAIL.  Can you ensure that?  Oh yeah, and did anyone mention that some places are already at their peak ventilator demand now and others will be there in weeks, so you have to sort all this out by yesterday.

Again, garage hacker is going to fail here even if they have 100% of the manufacturing design files.  Even big companies that make products with related processes is going to fail at immediately recreating someone else's design before they go through all the manufacturing process iteration.
Making a medical device has a seriously high bar for a reason!  The consequences of failure at any step here is that you may kill someone who otherwise might not have died.  That's not helping.  Any medical professional that has to debug your ventilator in the field is wasting time that could be saving someone's life.

Like Dave said.  A big chunk of this is mechanical, and he's not a mechanical guy, so this isn't what he does.  It would be so much easier if all the garage hackers could come to the same conclusion and stop this "lets build a ventilator" stuff. 

If you actually want to help now, go volunteer delivering groceries to some old folks. 

[G7PSK]

It would seem that CPAP machines would be better that the ventilators as they save people from needing the ventilators in the first place.
https://news.sky.com/story/coronavirus-mercedes-helps-develop-breathing-aid-as-f1-teams-join-ventilators-effort-11965584

[Smokey]

It would seem that CPAP machines would be better that the ventilators as they save people from needing the ventilators in the first place.

I read normal CPAP machines contaminate the whole room with exhausted virus.

[G7PSK]

It would seem that CPAP machines would be better that the ventilators as they save people from needing the ventilators in the first place.

I read normal CPAP machines contaminate the whole room with exhausted virus.

But most corona virus patients are in an open ward with other corona virus patients. It is only when things get really bad they go to intensive care the CPAP machines keep at least 50% of patients out of intensive care which is a big win, the CPAP machines are cheaper and easier to manufacture as well.

[nardev]

I already posted this on another thread. Just wondering what those guys from Turkey made? Can't find any more info for now, looks like they made a HUGE team to build this very quickly.



The post:

Check this out: https://twitter.com/Selcuk/status/1244661929777532930
Looks like the team: https://twitter.com/Selcuk/status/1244663216313155587
I'm i'm not wrong the pictures from the process: https://twitter.com/Selcuk/status/1243539495401783320

Also the clip is here:

[ormaaj]

It is doubtful that flexible licensing was a consideration in designing the software stack for a product like this. They're probabaly licensing dozens of 3rd party libraries and running some proprietary RTOS.  All of the supporting code and software will need to be replaced with something that permits at least 3rd party redistribution and ideally modification. Custom code for the specific device will then need to be reworked or rewritten around that.

[Michael_P]

take a look

[thm_w]

Other discussion here: https://www.eevblog.com/forum/chat/medtronics-open-source-ventilator-bullshit-pr/?topicseen

For the code, there is the chance they want to do an internal audit before release. You wouldn't want to release code that has any hint of liability eg "//this line caused vent to crash once and the motor stopped".
I'm not sure if you have to submit code for review on medical devices normally? If so then no excuse there.

[bushjavier]

Please check the website again they are releasing more files, the say more to come. The new release come with a lot of mechanical stuff.

Medtronic PB560 Ventilator System – Release 2.0 (.zip)
Includes:

Manufacturing fixtures
Printed circuit board drawings (including multiple BOMs)
3D CAD files
Mechanical part drawings

[tom66]

They've released schematics, gerbers, PCB layouts and BOMs, which actually makes these files quite useful now.  All that's missing is software which arguably is a huge part.  It would be nice if they at least released binary blobs for firmware, they could watermark the units with the "COVID-19 disclaimer" in the licence document if needed.

[BrianHG]

Please check the website again they are releasing more files, the say more to come. The new release come with a lot of mechanical stuff.

Medtronic PB560 Ventilator System – Release 2.0 (.zip)
Includes:

Manufacturing fixtures
Printed circuit board drawings (including multiple BOMs)
3D CAD files
Mechanical part drawings

Careful, read the required 'Permissive License'.  You need to register to get the cad files because it's designed to be a legal trap once the terms in the permissive license has reached it's imposed 2 possible deadlines:
https://youtu.be/TltM74rNasM?t=152

[EEVblog]

[Brumby]

I can see some possible reasoning behind the staged release of information...

Stage 1.  Paint a picture of just how involved the design of such a product is.  Give those opportunists who want to jump in to make a quick buck something to think about.

Stage 2.  Provide all the information to produce the physical product which will show just how much effort will be required to produce this thing.  If anyone wants to take this on, then they will need to be pretty capable!

Stage 3.  Release the software.  If you've made it this far, then why not be able to get the product out.


As for the licence - why all the cynicism?  Why all the criticism?  It is very clear Medtronics have invested quite a lot into the development of this product ... just imagine the hours that have gone into the documentation presented, let alone all the R&D that went into their production.  Why should they completely give that up?

As I see it, the licence is meant to make their design available for others to manufacture during a time of need and that once that period has passed, they are closing that offer and regaining direct financial interest.  They were entitled to this before the pandemic and nobody would have questioned it then, so why should their desire to reclaim this control be a problem?  Remember, the purpose of this action is the facilitation of health services - and this licence allows such an outcome.  It would be absurd to think that they are just "giving it all away".  Also, by spelling all this up front, no manufacturer should expect anything less ... and it allows them to make sure they are prepared for the inevitable end of such an arrangement.

Having the licence conditional on gaining the relevant approvals is just a further protection that the clones produced will be of an appropriate quality as to serve the purpose.  This prevents half-assed efforts from being produced and flogged off - which could cause more problems than exist already.



To put it another way - How would you present a design of yours for alternate production in a time of need without losing all benefits of something you spent years developing?

[ormaaj]

Why should they completely give that up?

I don't know that they're necessarily giving up anything. Anybody that wants is still free to buy one from them, which will be the vastly cheaper easier route. Major well-funded competitors won't give a crap or bother looking at this design. The only risk is in being undercut by Chinese knockoffs. U.S. health providers probably wouldn't touch something like that if it risks liability.

[Doctorandus_P]

When I saw the first blab, the first thing I thought was Jeez, give this company some slack.
It's probably a rush job in an understaffed company that may be trying to help while the phone is busy because everybody is trying to buy ventilators.

I gather it's a 10+ year old design. and it seems "normal" that design files "get lost".
They may have well have a software team working on figureing out if they are even allowed to release the sourcecode and the libraries to build the thing.
It's probably not their most sophisticated design. I haven't checked but it may have been out of production for years and they have to go through closets and old backups to find info.

This also seems to me to be the most logical reason for stuff being released in stages. They just release what they have when they've found it.

[ejtagle]

There you have: The 3rd link includes all the source code of both boards. BTW: I am not related to MedTronic in any way. I just tried a little bit harder to get all the parts of the project!

https://www.medtronic.com/content/dam/medtronic-com/global/Corporate/covid19/downloads/ventilator-files/medtronic-pb560-ventilator-system-file-set-1.zip
https://www.medtronic.com/content/dam/medtronic-com/global/Corporate/covid19/downloads/ventilator-files/medtronic-pb560-ventilator-system-file-set-2.zip
https://www.medtronic.com/content/dam/medtronic-com/global/Corporate/covid19/downloads/ventilator-files/medtronic-pb560-ventilator-system-file-set-3.zip