Last year, COVID-19 caused many of us to take a step back, but for some it was a chance to take a step forward.
Matt Carney, a recent PhD graduate from the MIT Media Lab’s Biomechatronics Group, was building a company that was trying to make bionic legs more accessible. When he went into lockdown, Carney found himself looking for ways that he could use his engineering knowledge to help.
“I found myself on the ‘Helpful Engineering’ Slack channel, which was basically an instantaneously generated, open hardware, global hackathon,” explained Carney. “The entire world was suddenly staying at home and, rather than read the scary news, around 13,000 people on that Slack channel began working ourselves into a frenzy.
“Face shields looked too easy, ventilators looked really hard and dangerous, but masks seemed solvable. At the same time, I was concerned about the safety of all the 3D-printed designs, and before I knew it, I was co-leading a team of strangers who were hell-bent on fixing masks.”
One of the main problems that Carney and his colleagues identified was traditional supply chains, which were unable to scale up production quickly enough to meet the surge in demand at the beginning of the pandemic.
Subsequently, distributed production methods such as additive manufacturing were utilized to supply front-line medical workers with personal protective equipment (PPE) like face masks and shields. However, Carney says that additive manufacturing technologies lacked a widespread level of material, production quality and sophistication, which limited the safety, repeatability and scalability for manufacturing N95 filtering facepiece respirators.
“We set out to build a new type of elastomeric, half-face respirator designed specifically to address the issues of the pandemic,” Carney continued. “High quality, scalable manufacturing, safe for health care workers, filter media agnostic and supply chain resilient. Those were the goals that we set out to achieve.”
The team focused on injection moulding, which Carney says leverages the best properties of elastomeric design—soft materials against the face, with rigid materials to support the filter.
With the number of designs coming in thick and fast, Carney soon found himself rebuilding the group’s document control structure almost every day: He first saved them onto his desktop, then into folders, then into more organized folders and shared folders.
“There were probably about 50 people from all over the world pumping out design iterations from nearly every type of CAD—SOLIDWORKS, Fusion, CATIA, NX, even OpenSCAD,” recalled Carney. To move forward, we needed our prototypes to be iterative and communicable across the team. What drove me crazy was there was not one native data set. So, I called Suchit at SOLIDWORKS, who I knew, and asked him for help.”
The problem Carney posed to Suchit Jain, Dassault Systèmes SOLIDWORKS Corp’s VP of Strategy and Business Development, was that he had a group of designers from different business units, different universities and different home labs all working on this one project and he needed to get them all on the same CAD system. He also needed document control that didn’t require setting up his own server or virtual private network (VPN).
“I talked this through with Suchit, who got me a handful of licences to distribute to the team. He also got us onto the 3DEXPERIENCE platform so we could do document control without setting up servers. It was amazing how fast this happened,” Carney added.
“All of this happened probably within the first week of the effort because, I think, SOLIDWORKS was also chomping at the bit to help. As engineers, we don’t sit idle while there’s a pandemic. We try to help. I could tell that was true because every call I got from every person was just so amped and gunning for a solution and trying to figure out how we could all help to solve this.”
Jain agreed: “We provided several online training/on boarding sessions to the ‘Helpful Engineers’ team working on this project. We also consulted and conducted fluid flow simulation on the mask within SOLIDWORKS. The flow simulation was done for NIOSH and FDA certifications. For example, one of the flow tests was to study accumulation of CO2 within the mask over a period of breathing cycles.”
Once the group members had a shared CAD tool, they were able to launch into design. One group set to work prototyping a design that they already knew had problems, aiming to learn why it didn’t work. A second group began working on a different design to move the project forward.
Carney said: “For the development of the mask, we made use of Simulation for structural, computational fluid dynamics, and mold flow. We use a master modeling approach to our design to ensure the fit between components, and, of course, also for document control.”
Designing the face piece and its semi-organic shape proved difficult. Carney explained: “Strangely, the design of the grill really stumped us. It is such a prominent feature, there were lot of strong opinions about it. It wasn’t so much the engineering as the design and user interaction that was difficult. From a modeling point of view, there are some tricky features about how to lock the components together that required some helical sweeps that proved difficult to generalize and be able to apply across multiple bodies to ensure that interlocking features fit.”
The founders of Formlabs, who Carney also knew, helped the group with printing and materials. “They have the world’s first liquid silicon resin 3D printer—Rapid Liquid Printing—and were able to print the 3D face pieces so we could test the seal and form of the mask as we were going,” Carney added.
After the design was finalized, the group then started testing: Breathing tests took place at ATOR Labs in Florida, which specializes in evaluating the breathing and metabolic performance of respiratory protective devices. Filter testing was conducted at the Rutledge lab at MIT. They also had clinicians on hand at Wake Forest University and Medical Center to conduct hands-on testing of their prototypes.
Dassault Systèmes helped with the computational fluid dynamics (CFD) simulations to evaluate air flow in the mask. This helped Carney and his group to reduce the size of the face piece. Over the summer of 2020, the group found a filter that worked for their design and also passed all the mechanical tests, filtration efficiency tests, breathing resistance tests and residual CO2 tests.
Carney said that the biggest challenges were not necessarily design or engineering related. “We effectively settled our design within the first six weeks—all of the time since then has been about spinning out a company, production tooling, and a quality management system,” he explained.
Once the design had passed these tests, Carney and the core group members created a nonprofit called Open Standard Respirator in order to distribute the product. “Then we learned that nonprofits aren’t great for hardware development,” Carney said. “Finally, somebody said ‘How about investment instead of donation?’ So, we created Open Standard Industries in order for us to scale the production of the mask.”
The Open Standard Respirator went on sale in December 2020 and the company is building an ISO-compliant quality system so that it can apply for a U.S. National Institute for Occupational Safety and Health (NIOSH) air filtration rating classification for filtering respirators. The ratings describe the ability of the device to protect the wearer from dust and liquid droplets in the air.
“Again, we’re relying on Dassault Systèmes for document control and quality assurance systems so we can keep everything together and controlled for our applications, because that’s the next step,” said Carney. “It’s not just passing a test—you have to have all of the quality management systems in place in order to get certified to guarantee that everything you’re making passes all the tests and you have statistical assurance of that.”
Open Standard Respirator’s goal is to go from zero to a certified mask within one year with its reusable mask with a tight, comfortable seal, high efficiency filtration (down to the size of the virus) and supply chain resiliency that filters in both directions.