Reliability-Based Hybrid Data Fusion Method for Adaptive Location Estimation in Construction

Abstract

Materials tracking and locating, which can be accomplished through various technologies and data sources, are key elements affecting construction productivity. The need for developing fundamental methods to take advantage of the relative strengths of each technology and data source while dealing with their limitations motivates the development in this paper of data fusion methods for improving materials location estimation. Particular attention is paid to situations in a construction environment in which radio-frequency identification (RFID) tags are attached to each piece of material, and the materials may be repeatedly moved around the site. The construction dynamics, the high noise ratio, and the limitations of the utilized sensing systems result in imperfect data that is imprecise and uncertain. A key challenge is using this imperfect data to improve accuracy and precision while maintaining cost-effectiveness and scalability. To address this issue, a hybrid data-fusion method was developed to increase confidence, accuracy and precision, and add robustness to measurement estimates. This hybrid method leverages evidential belief reasoning and soft computing techniques. The experimental results show that the hybrid fusion method outperforms the traditional methods in data fusion for location estimation. This study has successfully addressed the challenges of fusing data from a range of simple to complex sensor sources within a very noisy and dynamic construction environment. The results presented in this paper indicate that the proposed method has the potential to improve location estimation and to be robust to measurement noise and future advances in technology.