Fire and rescue stations need to plan predictively to always have enough personnel available. The expected volume of incoming calls is an essential factor in this process – often responders may experience increased volumes, due to several calls for the same event. But how we estimate the number of people dialing 911 at the same time?
In this case study, we will show you, how our software Premergency, makes prognoses for the workload of central, with stochastical methods and Machine Learning.
Predicting call volumes is a complex procedure, as many different factors may influence these volumes. More obvious factors include time and day. Moreover, factors such as the weather or events may influence call volumes significantly.
Personnel approach and planning solutions
In other words – the mean value of the respective time of a call is not enough to derive realistic prognoses about the number of incoming calls. Recognizing spontaneous peaks or overloads on specific weekdays is already a big advantage for emergency operators. Nevertheless, there is still a risk of lacking personnel, if external factors and possible correlations are disregarded.
In order to guarantee a secure and predictive planning process, we made to include as many indicators as possible, when developing this feature of Premergency. We visualize these indicators with the help of open data from the city of Seattle.
Based on historical data of the city and Seattle 911, we were able to identify influencing factors for our model. After that, we trained a model for the prognosis of incoming emergency calls. Furthermore, it is possible to predict temporal and seasonal peaks of the call volume. This algorithm marks the foundation for the latest feature of Premergency – the automatic prognosis of call volumes.
Planning Security for Fire and Rescue Stations
With this case study, we have shown, how to predict the number of incoming emergency calls using Premergency. Thanks to this prognosis, operators are able to prepare for peaks. This way, you have planning security and avoid overloading fire and rescue stations.