## The problem
A respirator is only as good as its fit on a specific face and the wearer's ability to don it correctly each time. Today, fit verification requires a $16-23K PortaCount and a trained operator. Daily seal checks rely on a "user seal check" with documented poor sensitivity (workers consistently feel they have a good seal when they don't). Matching face shape to respirator model still uses anthropometric data from 1960s US Air Force pilots, which leaves women, children, and non-white ethnicities under-represented. NIOSH's grand-prize-winning "Fit & Breathe" app from Mobomo was announced in October 2024. As of May 2026, 18 months later, it still isn't available to download. Mobomo's earlier app, FitMask, is on the App Store but only measures the face and maps to NIOSH respirator size categories. It does not check seal or detect leaks.
In a worst-case respiratory pandemic, lockdowns hold only if critical workers in energy, water, food and healthcare keep showing up. They will only keep showing up if they trust their respirator is protecting them and their family. A fit-and-don app, sitting on every critical worker's phone, is the difference between a stockpile that works and a stockpile that doesn't.
## What winning looks like
By 2028, a free open-source app, available on iOS and high-end Android, that:
- Generates a 3D face scan and recommends which respirators in the user's country stockpile are likely to fit their face
- Performs a quantitative fit test on the respirator they're wearing, with fit-factor accuracy validated against PortaCount in lab studies (target: ±20% agreement)
- Runs a 30-second pre-shift seal check that detects donning errors with sensitivity meaningfully better than the current user-seal-check
- Works for both disposable filtering facepieces (3M Aura class) and reusable elastomeric respirators (3M 6000, MSA Advantage 290, ProEquip's next-generation designs)
- Is endorsed in NIOSH guidance and adopted by 3+ national stockpile programmes (UK NHS, Taiwan CDC, EU rescEU partners)
- Has 100K+ active critical-worker users and a healthy open-source contributor base
The app is the missing software layer of the respirator stack. The hardware exists. The stockpile work is happening (slowly). Without this layer, the rest of the stack underperforms.
## Why now
- Mobomo's NIOSH Phase 3 win and the open-access Scientific Reports paper ("Thermal imaging and deep learning-based fit-checking for respiratory protection", 2024) have de-risked the underlying method. The IP is in the public domain. The remaining work is product, distribution and validation.
- Mobomo's failure to ship Fit & Breathe to the public 18 months after winning has left the lane wide open. Nobody is filling it.
- Smartphone hardware has caught up. iPhone Pro models since iPhone 12 Pro ship with LiDAR. The iPhone 15 Pro and 16 Pro have improved depth-mapping and on-device ML. A clip-on thermal-camera accessory (FLIR ONE Pro, sub-$500) gives any modern iPhone the input channel needed for the published leak-detection method.
- The next generation of reusable respirators is starting to ship. ProEquip, the Berkeley nonprofit working on civilization-scale biological crisis preparedness, is among the groups pushing on hardware. The software layer hasn't kept pace.
- AI-enabled biorisk concerns are pulling philanthropic capital toward critical-worker protection in a way that didn't exist three years ago. Coefficient Giving's biosecurity team expects to direct over $100M in 2026 alone, with explicit interest in PPE stockpiling and transmission-suppression technologies.
- The problem is solvable. It needs an entrepreneurial, tech-savvy founder to ship.
## What this could look like in practice
- **Open-source nonprofit.** A 4-6 person team (founder, two mobile engineers, one computer-vision/ML engineer, fractional respiratory-physiology advisor, partnerships lead). Ships a TestFlight beta within 9 months. Validation studies with NIOSH NPPTL and one national stockpile programme inside 18 months. v1 publicly available within 24 months.
- **Project absorbed into ProEquip or a sibling supply-side org.** Founder builds the app inside an organisation that already ships reusable respirators. Tighter coupling to one product line; faster path to distribution; less independent flexibility.
- **Research-to-deployment lab.** Founder partners with NIOSH NPPTL and a university computer-vision group on a 2-year programme: paper, open-source release, validation, then handoff to a deployment org. Slower; better academic credibility for regulator buy-in.
## Who should do this
- **Profile.** Has shipped a consumer mobile app at scale before. Strong taste for UX. Comfortable working with on-device computer vision and ML. Cares about pandemic preparedness deeply enough to spend three years on this. Comfortable building inside a nonprofit funding model rather than chasing venture economics.
- **Archetype.** Founder/operator energy. Not a researcher, though the team will need respiratory-physiology and computer-vision expertise.
- **Commitment.** Full-time, 3 years minimum.
## How to start
1. Email Mobomo and NIOSH NPPTL in week one. Find out exactly why Fit & Breathe never shipped. Buy a FLIR ONE Pro thermal accessory (around $400, USB-C version for iPhone 15 and later) and replicate the published method against a 3M Aura on your own face. Confirm it works on your hardware before pitching it to anyone.
2. Write a 6-page technical spec. Decide: ARKit Face Tracking vs ARCore for the 3D scan; built-in iPhone TrueDepth vs FLIR accessory vs stock-camera ML fallback for the thermal channel; which deep-learning approach you'll start from (the Scientific Reports paper is the obvious baseline). Identify the disposable-vs-reusable algorithm differences.
3. Run validation conversations with BlueDot's network in weeks two and three: ProEquip, Blueprint Biosecurity (Victoria Slaughter), Johns Hopkins CHS (Eric Toner, Anita Cicero), Coefficient Giving's biosecurity team. Confirm the open-source nonprofit shape is fundable, not duplicating live efforts, and that ProEquip would plausibly absorb the project at maturity.
## Source and endorsements
- **Field strategy:** BlueDot's Critical Worker Respiratory Protection field strategy (Section 5.2, "Sub-$500 fit-test device" and the train-the-trainer pipeline).
- **Written by:** Dewi Erwan, BlueDot Impact.
- **Currently pursued by:** Mobomo (Fit & Breathe, NIOSH Phase 3 1st place October 2024, $75K prize, not yet downloadable; their earlier FitMask app is on the App Store but only does face measurement, no seal check). OpenAeros (NIOSH Phase 3 2nd place, $50K, open-source bench-grade quantitative fit tester, hardware not phone). Consequent Labs (NIOSH Phase 3 3rd place, $25K, continuous-fit pressure-sensor wearable). No one is shipping a downloadable phone app to critical workers.
## Get involved
- **Funding pathway.** Exploration / career-transition grant for months 1-6: £100K-£250K for founder plus technical co-founder. Scale to £600K-£1.5M annually for a 4-6 person team across years 1-3.
- **Advisors and introductions BlueDot can offer.** NIOSH NPPTL, Mobomo, ProEquip leadership, Johns Hopkins Center for Health Security, Blueprint Biosecurity, Coefficient Giving biosecurity team, Founders Pledge biosecurity research team.
- **Contact.** Dewi Erwan, [dewi@bluedot.org](mailto:dewi@bluedot.org).


A small nonprofit team builds and ships a free, open-source mobile app that lets any critical worker pick the respirator most likely to fit their face, verify the fit quantitatively when they first put it on, and run a 30-second seal check before each shift, closing the gap between a stockpiled respirator and actual respiratory protection during a pandemic.

Open-Source Respirator Fit and Don Check App