A novel datamining method has been needed because the IoT (Internet of things) is spread over various kinds of industrial fields. We therefore propose to apply a data mining method to tool catalog data-base to improve the manufacturing technologies with machine tools because it is considered to include a useful information derived from tool manufacturing technology as a big-data. In the present report, we look at MIC (Maximal Information Coefficient) as a novel processing method to search for a new knowledge in tool data data-base, and construct a hierarchical clustering method based on MIC as a data mining method. Comparing a predicting equation derived from the conventional catalog mining method based on a traditional statistics with one based on MIC processing method, we investigate a function of MIC in data mining for end milling conditions. As a result, it can be seen that a constructed method (MIC data mining method) makes it feasible efficiently to find out essential variables in the radius end mill database because a derived practical formula has less interactions than conventional one with keeping the same prediction accuracy.

The purpose of our research is to establish a new RE (Reverse Engineering) method which efficiently generates easy-to-use solid models from scanned data of casting. The resulting models have properties of various types of surfaces and are beautified, which means the machining surfaces are aligned, for downstream applications such as duplication with machining and redesign. The proposed RE method is based on the separation of casting and machining surfaces in order to realize efficient RE process. The method consists of four processes. First, the surfaces of the scanned mesh are separated into casting and machining surfaces by existing roughness estimation method. Then, screw threads, through holes, and rotational small features are recognized from the separated machining surfaces. Next, beautification is done to make machining planes and cylinders parallel, orthogonal, co-planar, and co-axial using estimated dominant axes. Finally, beautified solid models with surface properties of castings are generated using Boolean operation of solid models generated from each surface. In thi s paper, methods for second, third and final process are newly proposed, and their performances are shown through the experiments. The effectiveness of the proposed RE method is also confirmed by comparison of RE time by our method with the one by skilled worker.

The on-the-spot inspections after accident are based on information obtained from the brake marks left on the road and the damage situation of the surroundings. However, this method is affected by the situation before the accident and cannot estimate quantitative guessing. Moreover, technologies such as automatic braking system and lane keeping system, have been developed actively. In developing these technologies, vehicle trajectory is one of the most important data. In this paper, we propose a novel method of estimating vehicle trajectory by using drive recorder images and point cloud which is measured by Mobile Mapping System and LiDAR. In the evaluation experiments, we confirmed the effectiveness of our method by comparing the vehicle poses estimated with RTK-GPS and showing the average errors are within one meter.

Unbalanced vibration compensation using a tracking filter is applied to Active Magnetic Bearing (AMB). This filter extracts only unbalanced vibration from the input signal. Peak-of-Gain Control (PGC), Auto-Balancing System (ABS), and Phase-Variable Control have been reported as compensation methods using this tracking filter. Furthermore, in the previous study, we proposed a compensation method using uniaxial tracking filter that suppresses unbalanced vibration in the axial direction. However, the compensation method using the tracking filter has a specific value for the cut-off frequency of the low-pass filter (LPF) incorporated in the filter, but there is no technical background of the value. In addition, the uniaxial tracking filter is derived by tracing the signal path of the conventional tracking filter. Therefore, there is no theoretical background that the uniaxial tracking filter extracts only unbalanced vibration. In this paper, first, the tracking filter applied to the radial direction is interpreted by using both the fixed coordinate system and the rotary coordinate system. Based on this interpretation, a method for adjusting the cut-off frequency of LPF is proposed. Next, through interpreting by communication theory, it is clarified that the uniaxial tracking filter extracts only unbalanced vibrations. Furthermore, a method for adjusting the cut-off frequency of the LPF in the filter based on this interpretation is proposed.

In semiconductor manufacturing processes, precise temperature control of the wafer is required to obtain high process performance. For this purpose, the relationship between the contact surface roughness and the thermal contact conductance was evaluated by using an experimental apparatus to simulate contact between the wafer and the stage in vacuum. As a result, when the gas flow between the contact surfaces was a free molecular flow (K_n>10), the gas component of the thermal contact conductance was not affected by the contact surface roughness but was proportional to the gas pressure. Furthermore, as the contact surface roughness decreased, the free molecular flow state remained even under high gas pressure owing to a decrease in the clearance between the contact surfaces. Therefore, a robust contact surface with respect to a change over time in the surface roughness and having high thermal contact conductance was realized by reducing the contact surface roughness and setting a higher gas pressure in the free molecular flow region.