AIR SAFETY: Creating and deploying Artificial Intelligence driven preventions of aerosol borne Respiratory diseases (AIR) for staff and patient SAFETY

The number and velocity of aerosol droplets produced when people breathe, speak or cough varies dramatically between individuals. We have engineered a human shaped device that generates aerosol droplets to mimic human beings speaking and have tested it in various clinical environments. We then computationally simulated and experimentally confirmed the impact of different mitigations on aerosol droplet concentrations.
Whilst screens and fans may prevent direct transmission of virus laden aerosol droplets between people, room-based solutions are needed to keep air clean particularly when people spend a long time together in healthcare environments. Portable high efficiency particulate air (HEPA) filters are cheap and quiet to run and can effectively reduce particle concentrations in poorly ventilated hospital consulting rooms. In airconditioned rooms these filters are still beneficial, although less so and their location becomes increasingly important. The impact is heavily influenced by the airflows caused by the air conditioning which are very difficult to predict. In general it appears that filters near the source of the aerosol droplets are likely to have greater impact than those in other parts of the hospital space. Our experiments in wards suggest that judiciously placed HEPA filters may reduce the risk of nosocomial infection spreading between patients. We are building an artificial intelligence system that will allow users to define the optimum configurations of HEPA or other types of filters for their own healthcare spaces.