RES – Real-Time Pollen Insights

What it does

RES is a smart pollen sensor that provides real-time, local allergy risk data. It solves the problem of broad, outdated pollen forecasts by giving accurate, hyperlocal insights to help people avoid exposure and manage their allergies more effectively.

Your inspiration

The idea came from speaking with people who suffer from pollen allergies and struggle with limited, inaccurate forecasts. Many told us they feel powerless during allergy season and want clearer information to manage their day. We also learned from NHS and MET Office experts that current pollen monitoring is outdated, expensive, and limited in coverage. Seeing this gap, we were inspired to create a low-cost, real-time solution that could help people take control of their health and reduce allergy impact through better data.

How it works

RES is a compact device that monitors pollen levels in real time. It works by using a small vacuum to pull in surrounding air. Pollen and other particles are trapped on a greased microscope slide placed inside the device. Once captured, the slide is positioned under a built-in microscope with a camera. The camera takes high-resolution images of the particles. These images are then analysed using a machine learning model trained to recognise different types of pollen based on their shape, size, and surface patterns. This allows the device to count and identify pollen species accurately. The results are uploaded to a cloud platform, where users can access local pollen data through a simple digital interface. By placing these sensors in multiple locations, we can create a high-resolution pollen map across a city or region. This system removes the need for manual counting, making pollen monitoring cheaper, faster, and more accurate.

Design process

Our design process began in summer 2024 with research into the lived experiences of people with seasonal allergies. We spoke to over 15 stakeholders including allergy sufferers, a doctor from the NHS, researchers at the MET Office, and a private garden scientist. These conversations revealed a shared pain point: current pollen forecasts are too broad, outdated, and inaccessible for people who need real-time, local information. We explored existing pollen detection methods and found they were expensive, manual, and limited in coverage. Inspired by bees’ natural ability to collect pollen through electrostatic charge, we began prototyping a low-cost system to detect pollen automatically. Our first prototype used a small vacuum and greased slides to collect airborne particles. We then added a camera and microscope to image the slide. We trained a machine learning model to identify and count different types of pollen based on visual features. Through multiple testing rounds, we refined the airflow, improved the lighting system for better image clarity, and simplified the slide mechanism for easier maintenance. We’re now preparing to run real-world testing and further improve accuracy, reliability, and usability.

How it is different

RES is different because it brings real-time, species-level pollen detection to everyday environments at a fraction of the cost of traditional systems. Current methods rely on large, expensive machines and manual counting under a microscope, with only a handful of sensors across the UK. RES automates this process using a compact, low-cost device with a built-in microscope and machine learning, making it possible to deploy in many more locations. Unlike other systems that only estimate overall pollen levels or use size-based filtering, RES can identify specific pollen types—essential for people allergic to certain species but not others. It also focuses on hyperlocal data, allowing users to see what’s happening in their immediate area, not just broad regional forecasts. This makes RES not only more accurate and accessible, but also more empowering for individuals and communities managing allergies daily.

Future plans

Our next step is to test RES in real-world outdoor environments to improve accuracy and reliability. We aim to refine the hardware for long-term use and scale up our machine learning model to detect more pollen types. On the business side, we plan to pilot with local councils, private gardens, and healthcare organisations. Long term, we want to build a nationwide sensor network to offer real-time pollen maps across the UK. Our goal is to empower allergy sufferers with better information, reduce health impacts, and support public health planning through affordable, scalable technology.

Awards

Our project won the Analytics for Society Award, recognising its potential to create meaningful social impact through data-driven innovation.