Respire

What it does

Respire tackles forgotten and wasteful air filter changes by using machine learning to predict when filters need help. It automatically sprays a corn husk-based coating, restoring performance and extending filter life for cleaner, greener air.

Your inspiration

As wildfires filled my city with smoke last summer, I started thinking about the importance of clean air. Looking at all the plastic single-use air filters, I knew I wanted a sustainable option. Being from Canada, corn husks became our starting point. After months spent developing a prototype for areas hit by wildfires and bad air, I thought of myself. I hadn’t changed my filter in so long. We needed a fast solution to make filters sustainable and clean air accessible. That’s how we arrived at Respire: a spray to renew filters and make clean air easier for everyone.

How it works

Respire is a simple spray system designed to make air filters last longer and work better. When air filters get clogged, they can no longer trap harmful particles, so most people throw them away and buy new ones. Respire changes that by using a spray made from corn husks, a common agricultural byproduct in Canada. Sensors are placed near the air filter to monitor tiny particles in the air (PM2.5), as well as temperature and humidity. This data is used to train an AI model to predict when the filter is about to become ineffective. At that point, Respire will automatically spray a fine mist of a specially prepared cellulose nanofibril (CNF) solution, made from corn husks, onto the filter. This forms a thin, invisible layer that helps the filter trap more dirt and particles, extending its usable life. This approach saves money, reduces waste, and uses renewable materials to make clean air more accessible and sustainable for everyone.

Design process

We began by extracting cellulose nanocrystals (CNCs) from corn husks and freeze-drying them to create a lightweight aerogel foam. While this material was extremely porous, we quickly realized it was too brittle and hard to shape for practical use in standard air filters. As we dug deeper into the research, we found that cellulose nanofibrils (CNFs) offered better flexibility and the ability to form a web-like layer, which literature suggested was more effective for filtration. We shifted our focus to extracting CNFs instead of CNCs and started brainstorming how best to implement them in real-world filters. Shaping foams or integrating sheets remained a challenge, so we explored the concept of a spray that could coat existing filters with a nanofibril layer. The next question was when and how to apply the spray for maximum effectiveness. Currently, we are developing a system that uses machine learning and sensor data to monitor filter performance and automatically trigger the spray at the optimal time. This approach promises a smarter, more sustainable solution, and our next steps are to build and test functional prototypes of this new model.

How it is different

Respire stands out from existing air filter solutions in several ways. Most products on the market, such as Filtrete Smart Filters and other sensor-based systems, only remind users when it’s time to replace a disposable filter. Some advanced filters use synthetic or cellulose fibers, but these are designed for single use or require mechanical cleaning. Even the newest bio-based filter research focuses on rigid membranes or foams, not on-site regeneration. Respire is the first to combine a spray-on, plant-based nanofibril coating (made from waste) with machine learning and real-time sensors to automatically renew filters. This approach not only extends filter life and reduces waste but also works with existing filters, making it retrofit-friendly. No current competitor uses smart automation and a renewable, sustainable spray to refresh filters in place, making Respire a unique step forward in both air quality and environmental design.

Future plans

We’re currently collecting in-home data to train our machine learning model for optimal spray timing. We have a CNC gel-like mixture but our next step will be to extract CNFs directly to align with the latest literature, and then test the spray’s filter capacity in the lab and in real-world settings. Once validated, we aim to deploy a standalone unit for any HVAC system. Our original goal is to provide air filtration in emergency situations such as remote clinics and wildfires so also plan on testing our filter in simulated emergency settings and on filters well past their breaking point to see how much we can salvage.

Awards

We were finalists at the Biodesign Challenge for our initial corn-based CNC aerogel air filter concept, before transitioning to our current spray method.