Utilização de dados de ocorrências para tomada de decisão no Corpo de Bombeiros Militar do Distrito Federal

Abstract

Com a crescente quantidade de informações que permeia a vida de todos, é essencial que se adotem métodos objetivos de utilização e analise de informações. Este trabalho tem como objetivo demonstrar a importância do uso de dados para a tomada de decisão, pelo confronto entre as percepções pessoais dos bombeiros e a análise de dados. Para verificação das ideias existentes entre os bombeiros sobre os fatores predisponentes ou temporais dos diversos tipos de ocorrências no CBMDF, foi realizado um questionário com oficiais do quadro QOBM/Comb. Os dados de ocorrências foram extraídos do sistema Fênix e analisados sob a abordagem da estatística inferencial. Apesar de não haver tendência de aumento no número de ocorrências como um todo próximo aos dias de pagamento, como esperado pelos bombeiros, verificou-se aumento no número de ocorrências de APH. A hipótese de um maior número de ocorrências envolvendo motocicletas nas primeiras horas da manhã, segundo a percepção dos oficiais, foi confirmada pelo modelo escolhido. Ha, na Corporação, grande desconhecimento acerca dos padrões temporais de ocorrências de tentativas de suicídio. Um modelo exponencial foi eficiente para explicar boa parte da variação no número de ocorrências de incêndios florestais em função da umidade. Este trabalho demonstrou que informações empíricas não necessariamente refletem o cenário geral, algo que só pode ser verificado pela análise dos dados. Fica evidente, ainda, o desconhecimento na corporação acerca dos padrões de certos tipos de ocorrências. Os dados resultados deste estudos ressaltam a aplicação e a importância da tomada de decisão baseada em dados.

With the growing amount of information in everyone’s lifes, it is essential to adopt objective methods of using and analyzing information. This work aims to demonstrate the importance of data-driven decision making, by confronting the personal perceptions of firefighters and analyzing data. To check the existing ideas among firefighters about the predisposing or temporal factors of the different types of calls in the CBMDF, a survey was conducted with officers responsible for main decisions. Emergency calls data were extracted from the Fênix system and analyzed under an inferential statistics approach. Although there is no clear increment on the number of occurrences as a whole close to paydays, as expected by the firefighters, there was an increase in the number of pre-hospital care calls. The hypothesis of a greater number of calls involving motorcycles in the early hours of the morning, according to the perception of the officers, was confirmed by the model chosen. There is, in the Corporation, great lack of knowledge about the temporal patterns of suicide attempts calls. An exponential model was efficient to explain much of the variation in the number of occurrences of wildland fires as a function of humidity. This study showed that empirical information does not necessarily reflect the general scenario, something that can only be verified by analyzing the data. It is also evident that CBMDF is not totally aware of the patterns of certain types of occurrences. The data shown here underscore the application and the importance of data-driven decision making.

With the growing amount of information in everyone’s lifes, it is essential to adopt objective methods of using and analyzing information. This work aims to demonstrate the importance of data-driven decision making, by confronting the personal perceptions of firefighters and analyzing data. To check the existing ideas among firefighters about the predisposing or temporal factors of the different types of calls in the CBMDF, a survey was conducted with officers responsible for main decisions. Emergency calls data were extracted from the Fênix system and analyzed under an inferential statistics approach. Although there is no clear increment on the number of occurrences as a whole close to paydays, as expected by the firefighters, there was an increase in the number of pre-hospital care calls. The hypothesis of a greater number of calls involving motorcycles in the early hours of the morning, according to the perception of the officers, was confirmed by the model chosen. There is, in the Corporation, great lack of knowledge about the temporal patterns of suicide attempts calls. An exponential model was efficient to explain much of the variation in the number of occurrences of wildland fires as a function of humidity. This study showed that empirical information does not necessarily reflect the general scenario, something that can only be verified by analyzing the data. It is also evident that CBMDF is not totally aware of the patterns of certain types of occurrences. The data shown here underscore the application and the importance of data-driven decision making.