A team at the University of Minnesota and the partner companies designing the components of a mechanical ventilator have pulled off a remarkable feat: Just three weeks since work on the design first began, pre-production units of the mechanical actuator for the ventilator are now shipping and full production has just now started.
The device, called the Coventor, is based on a manually operated bag-valve mask (known as a BVM or Ambo bag) design, but uses an electric motor/actuator to automatically squeeze the bag. It isn’t designed to replace an ICU ventilator, but rather to provide health care institutions a life-saving option when standard ventilators are needed but no longer available. With a starting price of around $1,500, the expectation is that in volume it could drop to well below $1,000.
Orders for 2,000 units have come in and pending the FDA’s granting of an Emergency Use Authorization (EUA), several thousand ventilators can be manufactured and shipped in a matter of weeks. The use of off-the-shelf parts mean hundreds of thousands of components could be sourced quickly, meaning even higher volumes.
Below is a video of the first prototype in action. The full playlist here links to a series of videos chronicling the progress of the project.
How this quick action was even possible required a complete upending of the traditional mindset of many of the engineers involved.
“We do a lot of work with medical device companies and a typical product from conception to the point it’s ready for FDA approval might take four to six years,” said Jon Klipfel, a senior sales engineer at Teknic, which developed the motor design. “In this case, we needed to quickly iterate the conception-prototype-test and validate process with the goal of high-volume, rapid deployment in about two weeks."
And speed wasn’t the only harrowing requirement. Teknic, which designs high-precision servo motor systems used in many medical applications, needed to eliminate electronics and software wherever possible to minimize the complexity of the design.
“High precision motion control is our forte, and what was so different about this project was the unrelenting simplicity,” stressed Klipfel. “We wanted this to be open source, so this wasn’t a design where we wanted to use sophisticated motion control algorithms.”
Electronics distributor Digi-Key, who collaborated with the University to create a plan for the ventilator and source parts, initially recommended Teknic, known for its high-quality products, to work on the motor/actuator design—the primary component of the design. “We met with the team and it kind of took off from there,” said Klipfel. “We put aside anything that wasn’t 100% essential at that time to focus on the motor/actuator design.”
The final design would need to be a simple motor connected to some mechanics and a power supply to control the bus power to the motor.
“We knew this design had to run off of a readily available 24 power supply and we used a simple encoder knob as a method of turning up the speed, which is directly proportional to the respiration rate,” he explained. “It’s a last resort type of thing, but we knew you could get the parts in super high volumes.”
To get feedback on their efforts, the Teknic team turned to social media, posting videos on YouTube and using Reddit to solicit comments.
“As far as the social media aspect of this whole thing, this was really our first foray into it,” noted Klipfel. The idea was to create some awareness about the design that we were working on and hopefully to spark some ideas, and we definitely achieved that.”
Once the initial videos and posts were created, Klipfel said the information and call for comments started popping up all over the place online. “There was some very useful feedback, especially around an early design that had a couple of spur gears in place [to accommodate the use of alternate motors around the design], with engineers weighing in on using it as a direct drive instead.”
Not surprisingly, given a community of engineers, some of the commentary centered on ways to do things that involved advanced, fancy features. “People were questioning why we didn’t try this or try that, and there were some really creative ideas for sure,” said Klipfel. “But at the end of the day, one of our critical goals was to be able to scale extremely quickly and some of those suggestions were counter to that goal.”
With the hyper-compressed schedule Teknic engineers had created a solid model of a spinning gear with a motor in 24 hours, and less than a day later they were 3D printing parts. “For the first prototypes we did not have bench metal, we had to work with what was available,” said Klipfel. “So, one of our engineers just grabbed some a spare four-by-four and plywood; it was that level of engineering."
To date, testing of the motor/actuator is going well and a limited number of pre-production units have already shipped. Full production has just begun. Once the unit passes a minimum set of tests, the University of Minnesota will release the final design files.
Once the FDA grants a suitable EUA, Teknic and its partners will be able to quickly build several thousand units. While these early units are being built, engineers will modify the design to use a simple DC gearmotor and controls (designed, but not sold by Teknic) that can be swiftly sourced by the hundreds of thousands.