Drone assisted emergency response model for Scotland

A drone-based network to assist health services has been an emerging topic in recent times. Many countries are already exploring the use of Drone based logistic network for medical goods delivery. Following this path, UK government is also exploring this idea through project CAELUS (Care & Equity – Healthcare Logistics UAS Scotland) consortium. A part of this service is to use drone logistic network for supplying Automated External Defibrillators (AEDs) to assist patients suffering from cardiac arrest. This is particularly important as even though it is mandatory to send an ambulance to a cardiac arrest call, it is not always possible to reach a patient in time because of certain external influences such as availability of an ambulance, traffic conditions. A drone network will therefore allow us to send an AED as early as possible to the patient so that they can be revived in most of the cases by the time an ambulance reaches. Having this in mind, we construct a facility location problem that aims to improve the resilience of the emergency response service provided by Scottish Ambulance Service (SAS). To do so, we tackle this problem as a time dependent problem, where we want to achieve a certain probabilistic threshold of drone coverage by allocating drone stations across different sub-regions of Scotland. This region-specific problem allows us to solve the optimisation problem with lesser computational term but also serves a very important aspect of the emergency response system, ‘healthcare for all’. Any sort of statistical model relies on observational data and therefore we need to be very careful of the heterogeneity present in the data. Since, the number of recorded calls is very much dependent on the population of the area, having a global probabilistic model will only benefit the patients from the metro cities where most people reside. However, a region-based approach will ensure that we have drone stations in different parts of Scotland which will serve the patients from rural areas. Based on this approach we present a case study based on the region called ‘Grampian’ and show different metrics associated with our analysis such as percentage of land covered, percentage of population settlements covered and most importantly expected percentage of cardiac calls that we can cover. Moreover, we compare the case, where we only use the ambulance stations as drone port to observe the metrics we mention before. This is particularly interesting as constructing a drone station comes with additional cost both in the design and operation phase. So, using the ambulance station also reduce those cost, which might be beneficial in certain cases.