Avaliação da precisão e exatidão na coleta de dados GNSS em dispositivos móveis em função do tempo de observação e plataforma de desenvolvimento

Abstract

The present study was motivated by the increasing use of smartphones in geolocation applications and the need to evaluate the influence of different development approaches on the accuracy of the data obtained. To this end, specific applications were developed for collecting GNSS data in a controlled environment, using Xiaomi Redmi Note 10S smartphones. The experiments were carried out in an area with fixed paddocks, with data collected at different times (morning and afternoon) and varying observation times. Statistical analysis of the data included the application of ANOVA and Tukey’s mean comparison tests, allowing the identification of significant differences between measurements. The results indicated that, statistically, the factors observation time and development platform were not significant, however, they indicated that the precision and accuracy of measurements is higher in the morning, with less dispersion of errors, while in the afternoon the errors are greater and more dispersed. Specifically, Java platform had a slightly lower average error rate in both the morning and afternoon compared to Flutter. The graphical analysis of the data dispersions showed that, during the morning, the errors are smaller and more concentrated around the reference point, while in the afternoon, the errors are more dispersed, indicating a decre- ase in measurement precision. Factors such as atmospheric conditions, satellite position and less human interference during the morning are highlighted as possible causes for the variations.