A machine learning approach to predict production time using real-time RFID data in industrialized building construction

Data- and Knowledge-Driven Cycle Time Estimation in Offsite Construction

“…In another study, NNs and multivariable linear regression (LR) models were applied to precast concrete production to estimate productivity in consideration of influencing factors related to product shape, material, and manpower [13,14]. Conducting research on productivity in wood panelised construction, Mohsen et al [15] trained a number of different machine-learning algorithms to estimate the time required to complete processes on a wall production line as a function of design-related factors (e.g., length, width, number of studs) and factors related to work in progress (e.g., the count of wall panels being processed on the production line). In another recent study in panelised construction, this one targeting the transportation phase, Ahn et al [16] trained support vector regression models using GPS data in order to predict transportation time for a given project as a function of product-related factors such as the total floor area and total wall area, as well as site-related factors such as location and the maturity of the neighborhood.…”

“…Various NNs have been developed to imitate desirable characteristics of the human brain such as its learning ability, generalization capability, and adaptivity [40]. In the aforementioned study by Mohsen et al [15], however, among the models considered-i.e., random forest (RF), LR, knearest neighbor, and NN-the LR model was found to perform slightly better than the NN model when trained on an engineered dataset to predict the production time of wall panels, and the best performing model was the RF model [15].…”

“…It is worth noting that, when the framing elements (i.e., studs, D-studs, M-studs, large windows, etc.) were combined into a single feature following the approach outlined in the study by Mohsen et al (2022) and used along with the remaining features in the model, the resulting error was higher compared to when including all 24 features, with the RMSE measured at 3.04 min and the MAE at 2.26 min.…”

Precast production scheduling in off-site construction: Mainstream contents and optimization perspective

“…Yan et al (2022) developed an intelligent monitoring and evaluation system for precast construction schedules based on computer vision-based (CVB) technology that could help production managers efficiently track the real-time status of PC and workers. With the advancement of information technology, the Internet of Things (IoT) technologies, such as RFID, have been applied in the OSC domain to facilitate the tracking of PCs' real-time status (Li et al, 2017;Mohsen et al, 2022), Luo et al (2020) collected and managed real-time information of PCs through an information platform combining building information modeling (BIM) and RFID, which was proven to help managers manage risks and adjust project schedules more efficiently through a case study.…”

Barriers to the Implementation of Radio Frequency Identification (RFID) for Sustainable Building in a Developing Economy

“…The introduction of RFIDs can boost the establishment of sustainable building partners. Applying the RFID method by building associates allows storage and access to stored data instantaneously [24,25].…”