Completely Automatized Recognition of the Installation Process in Special Foundation Construction

Up to today, it has been very difficult to obtain qualitative and quantitative information about the status and production process of a construction site. It is becoming more and more important to be able to analyze this process so as to notice bottlenecks and interruptions on the site early on. The high costs of each shift in special foundation industry mean that making the wrong decision and/or interruptions can put the financial result at risk. Previously, keeping reliable data meant manual process documentation. This was done either by an additional staff member (incurring high additional costs for training, education, etc.) or by the equipment operators. The latter is critical, because they have to initiate every process step change, leaving no time to log data separately. Studies have shown that there are considerable deviations from the actual process steps. In this paper, a seamless, automated method will be introduced to analyze the construction process. This method recognizes each process step from the equipment data. The main purpose of the data recorded by the piling rig is the need of quality reports and now basis for the presented analysis. Here, a method was successfully transferred from theoretical computer science (compiler construction field) to find application for process recognition. Because the execution of works in special foundation construction can always be defined sequentially and therefore also practiced sequentially, recognition can occur step-by-step. Our method always recognizes the next step, so it does not have to monitor as much data at the same time. Individual trigger points are easier to define. Unlike solutions that are based on a global approach and allocate suitable process steps to data patterns, thereby also creating gaps, the presented method is always seamless. Furthermore, this method can be applied regardless of time or location provided that the necessary machine data is available, which is normally the case with production data logging. This means that construction sites even from the past can also be analyzed and the results used for comparisons. The first results from the analysis of a complete continuous flight augering (CFA piles) installation process of a construction site will be presented.