As an engineer you tend to see the world as a bunch of problems; and how to split a ventilator is a fascinating one.
I'm going to go rogue and vote we put patients in SERIES, not parallel as everyone keeps suggesting.
I call this concept: ventilator gating.
Instead of delivering one large breath to 4 people; why not deliver them one at a time. Most ventilators allow for pretty high respiratory rates. Thus a single ventilator, if it can supply 60 breaths a minute to one patient; can in theory deliver 15 breaths a minute to 4 different patients. All the parameters, rules and alarms for a single patient can continue to apply without risk of over inflating one patient when their compliance or size is different than the others in the network.
The easiest example is two patients on one ventilator. In the method described from 2006 and everyone is leaping to initiate put both patients onto the same inhalation line, and same exhalation line. This works great if both patients have the same volume, same resistance/compliance, and same need for PEEP/volume/O2 concentration -- the greatest danger would arise if their compliance or volumes were different. Imagine inflating two balloons, one as it is very small and requires tremendous effort to inflate and one that is already half full -- all the air will preferentially go to the easier to inflate (and already partially full baloon) -- you will eventually POP one of those balloons.
Instead, give one breath to patient 1 and as they exhale, switch the ventilator to a second patient, and give them the second breath. If you could rapidly connect/disconnect the patients you would never have to worry about popping lungs, and time spent in exhalation for one patient won't mess with the machine trying to inflate someone else. This is especially useful when someone has a prolonged exhalation period.
So instead of this (the popular method):
There a a million mechanical and mechanical/electrical systems can can enable switching. I'm doing to build one of each to prove it can be done. I will then ask every mechanical engineer out there to build a better one!
The first mechanical system I developed is a lock and key method; exhalation directs airflow to the other patient. The top trio is for inhalation circuit (single intake of air, and can direct to 2 patients -- the key aims the air to patient 1 then patient 2 depending on its orientation. The bottom trio is where the exhausts meet up at the end -- the airflow of exhaust sets the sail straight, which redirects the upper section of the key to point at the other patient. No electricity needed!
A mechanical/electrica lsystem is easily realized with a Lego robotics kit. These babies are cheap, reliable, can be battery powered or plugged in -- for this I use a lock and key method again; but the trigger is a spring pressed up against a touch button sensor (standing in the lego kits); and program the lego motor (which has an encoder and motor combo -- so you can set rotations and direction using graphical programming -- I taught this stuff to 4th graders! ) So each breath delivered trigger the device to rotate after 1 second. For this to work the inspiratory time must be less than the delay.
There are prefabricated parts that do this; they are called solenoids. They are shut valves; that open when energy is applied to them, and a simple circuit could switch between the 2 solenoids and take turns opening them. The problem is, the are much more expensive, has very high resistance in their circuit, and require a lot of extra parts to mate them together to standard vent system.
If there was a commercial system, I suspect they would use these prefabricated parts and develop some hardware and a circuit to go with them. I don't think we have enough time for this type of system -- so my guy says we should look for a pure mechanical system. If we have to use electro/mechanical system then we need to use cheap parts that are available worldwide -- and Lego robotics hits this criteria. Thanks to Professor Chris Rogers so making it idiot proof -- even 4th graders can build and program these things graphically -- and they are highly reliable. I used to watch children program submarines on their laptop and launch them to collect items in a pool.
When I get a ventilator and print a clean mechanism I'll show you the model in action. I also ordered some test lungs.