Land subsidence, which used to be a serious social problem, is now under control due to groundwater pumping regulations enacted in the 1960s. However, as a result, the recovery of the groundwater level has been promoted, and the new problem of ground uplift has been occurring mainly in urban areas. However, the risk of ground uplift disasters is a concern. In this study, we apply PSInSAR, a type of time-series SAR analysis, to X-band imagery, which is well suited for analysis in urban areas, to monitor areal ground deformation. As a result of the analysis, the RMSE is 1.6 mm/year compared to the level survey data, indicating that the spatial interpolation of the level survey data is possible. Furthermore, a strong correlation (r = 0.80) was found between the amount of groundwater level rise and the amount of ground uplift.

 Although a shield tunneling machine should excavate a tunnel along its planned alignment, deviations occur between the planned alignment and the actual result. When this happens, it is necessary to draw a target alignment that returns deviations to the planned alignment gradually and to excavate with the shield machine along the target alignment. These operations are controlled by skilled engineers and trial-and-error operations, and labor-saving is an issue. In addition, skilled excavation managers and operators are aging, their skills may be lost in the near future. Therefore, machine learning is expected to play an important role in automating the operation of shield tunneling machines. However, related studies have been unable to automatically calculate the optimum operation parameters for excavation along the target alignment. Therefore, this study sought to use machine learning to develop an autopilot model, which is a method to automatically calculate optimal operation parameters of the shield machine for the straight section of a planned alignment.

 This study theoretically derived formulae defining the upper limit of precision in estimating three dimensional coordinates by UAV-based photogrammetry under simplified conditions. These formulae evaluate the horizontal and vertical variances of multi-view triangulation (the principle of coordinate estimations for both point cloud and validation points) as functions of flight-design variables such as image resolution, focal length, camera altitude, and overlapping ratios. They can be used for a preliminary check to avoid flight designs that will never meet the desired precision.

 This paper attempts to estimate the aggregate volume in an outdoor asphalt plant by using a single camera and Convolution Neural Network (CNN) based-image processing. An end-to-end prediction model between a captured image and a volume was used. However, its accuracy was heavily affected from sunlight. To solve this problem, a new data augmentation technique is employed for the effective training. This technique generates virtual training images with many variations of sunlights. Our CNN trained by this dataset has suppressed the estimation error to about 8%. The experimental results have shown that the proposed method is robust to the outdoor noises.