Risk

Working in the software industry as I do, I like to think that I recognise "risk" to do with the software development life-cycle where many others don't (which is, of course, true for experts in most industries) With everyone and his dog rushing to build ventilators (often for dubious reasons IMO) there is a distinct risk that the underlying control software in these brand-new designs and machines will have faults (In 35 years of development, I've never released a piece of software without a fault in it somewhere) Now, hopefully any faults turn out to be benign and not cause dangerous situations, but then again, that's the thing with faults, you simply don't know where they are (or you'd fix them!) and so with new minimally tested software the risk is much greater. 

Here's a story about a ventilator that in my view takes a slightly more pragmatic route, it's almost entirely manual using what we used to call a "belt and braces" approach. It's made by a team at MIT in the USA and is centred around an "Ambu bag", which is the blue thing in the picture and what's used to manually ventilate patients when fully automated machines aren't available (i.e. a doctor squeezes air in the bag into the patient by hand). In this prototype a simple motor squeezes the bag containing an air/oxygen mix via a tube inserted into the persons' mouth/throat. The other key thing about this design is that it costs around $100 in parts rather than the $30,000 that much more sophisticated and automated machines cost. So, although perhaps not the ultimate in medical technology, things like this could prove to be real life-savers once this thing gets into it's stride in India and Africa, there certainly are a lot of people out there who'll need something at the more "robust" end of the spectrum to help them get through this.