Artificial Intelligence for Pollen and Spore Detection, Forecasting and Human Health (AIPS)

Pollen and fungal spores are important for human health in both outdoor and indoor environments. They are linked to several respiratory illnesses which range in severity from minor to deadly. Better detection and forecasting of pollen and fungal spores would allow for interventions to be developed that would reduce their risk to human health. 

The current methodologies available for the detection of pollen and fungal spores are either expensive or time consuming, and often both. This hugely limits their use. For example, the UK Met Office currently only has available 11 regulatory grade sites for pollen monitoring from which their pollen forecast is based upon. This equates to about one regulatory pollen monitoring station per 11 million people in the UK. Similarly, regulatory agencies lack cheap methodologies to detect fungal spores in both outdoor and indoor locations. A cheaper, more agile detection method would much increase the UK's capacity for the detection and forecasting of pollen and fungal spores. 

The AIPS project has combine several rapidly developing technologies. It brings together a distributed internet-of-things (IoT) sensor arrays in combination with regulatory grade equipment and artificial intelligence (AI) techniques. The IoT sensors measure the size distribution of the small particles that are present within the air. The sources and compositions of these particles are many and varied. Atmospheric particles include bioaerosols that are composed of fragments from the biosphere, including pollen and fundal spores. Finding these bioaerosols within the much larger populations of other atmospheric aerosols, is like finding a needle in a haystack. Fortunately for this project, pollen and fungal spores have well defined sizes that are distinct to the background aerosol which makes detection possible. AI approaches will use machine learning algorithms to classify the pollen and fungal spore species of interest and generate approaches to detect them in real time. The results are then compared to regulatory grade equipment to assess the skill of the low-cost approach.  This real time detection will allow for data-driven real-time forecasts of the pollen and spore species of interest.

The presentation will provide an overview of the AIPS project, and discuss the efficacy and applicability of the new AI and IoT tools with respect to bioaerosol detection and forecasting needs.