In this paper, the development of CNN-based appearance inspection method named “asymmetric label smoothing method”, which reduce missing of defective products in appearance is reported. In the manufacturing site, it is necessary to inspect finished products and to remove defective products for quality assurance. The most important function required in appearance inspection is to minimize the number of defective products mistaken for normal ones. The proposed method proactively detects products with suspected defects, thereby reducing the number of missed defective products and over-detected of normal products. In the proposed method, asymmetry is given to normal and defective products, and the label smoothing method is applied among the classes of defective products. Using the model trained by the proposed method, we conducted an experiment of appearance defect detection in automobile clutch discs and succeeded in reducing the number of missed and over-detected compared to the conventional method. The proposed method is expected to be an effective method to improve the accuracy and to reduce the number of missed in defect detection at actual mass production processes.

Detecting the center of gravity (COG) is important to evaluate the rollover risk of a forklift because it lifts various kinds of cargo with unknown COG．Some conventional methods have been proposed to detect COG, however, they require the forklift to stop moving or long seconds of measuring data, Therefore, we propose a method to estimate COG in real time by measuring the relation between the acceleration and the tilt moment of the mast. The COG is found by grid search to satisfy the above relation for past several seconds. It does not need any constraints of forklift motion. This method is modified for the conditions of mast motion and body turning. We implemented the sensor system on an actual forklift and the effectiveness of our proposed method was shown by various experiments with it.

In femtosecond laser processing, it is difficult to accurately process a three-dimensional shape by 2.5-dimensional irradiation paths because a three-dimensional removal shape by ablation is not identified. Therefore, a femtosecond laser CAM system based on the identified three-dimensional removal shape is required to process a three-dimensional shape with curved surfaces and highly inclined surfaces. This paper proposes a novel method to generate femtosecond laser processing paths for processing the three-dimensional shape. In the method, a virtual tool shape of femtosecond laser processing is identified based on a removal shape formed by femtosecond laser irradiation. Femtosecond laser paths to process a required shape are generated based on the inverse offset method using the virtual tool. An unprocessed shape (undercut) by femtosecond laser processing was identified by a femtosecond laser processing simulation system and paths to process the unprocessed shape changed in the setting angle of a workpiece are additionally generated. The change in the setting angle of a workpiece is continued until there is no unprocessed shape. The effectiveness of the proposed method was then verified by experiments.