Development of a Sensor System for Agricultural Machines using Stereo Vision and Deep Learning

( Kosuke Inoue ),( Yutaka Kaizu ),( Kenji Imou ) 한국농업기계학회 2018 한국농업기계학회 학술발표논문집 Vol.23 No.1

Autonomous navigation of agricultural machines by employing the global navigation satellite system (GNSS) has developed rapidly in recent times. However, a machine that is only based on GNSS cannot detect obstacles such as humans, which will cause an increased risk of collision with such obstacles. Further, conventional distance sensors cannot accurately determine the distances to various obstacles because of grasses or crops that may lie between the sensors and obstacles, blocking the sight of the sensor. To overcome this problem, we have developed a sensor system that can precisely determine the distances from the sensors to humans even in the aforementioned circumstances. We combine human detection, based on deep learning, with distance detection by means of a stereo camera. Human detection with deep learning can be used to obtain an RGB image from the stereo camera in order to classify obstacles and to detect their locations in the image. When humans are detected, the detection image is compared with a depth image, and the location in the distance image is determined. Further, the median of the distance values corresponding to the pixels at the detected location is calculated. Using this sensor system, we measure the distances from the sensors to a human who was standing in a vegetated region. The errors are 2.2, 4.9, and 14.5 cm, respectively, for distances of 2, 3, and 4 m from the camera. The results depicted that this sensor system exhibits sufficient accuracy in case of agricultural machines.

Scheduling for Assembly Line with Human–Robot Collaboration

Kosuke Inoue,Hideki Aoyama 한국정밀공학회 2023 한국정밀공학회지 Vol.40 No.2

In the scheduling of assembly lines with human–robot collaboration, variations in workload caused by differences in the available working hours of workers and robots must be minimized. A scheduling method that considers buffers shared by automated guided vehicles and cooperative assembly by multiple workers is proposed herein. In particular, cooperative work requires an assembly schedule that minimizes the make span and satisfies the delivery date, while accounting for the possibility of work partitioning, the number of workers, as well as their available time slots and skills. Hence, it is difficult to obtain an exact optimal solution within a reasonable computation time using existing methods such as mathematical programming. Heuristic or metaheuristic approaches are effective for solving this problem. However, these approaches are not suitable for cooperative assembly by multiple workers. Therefore, a genetic algorithm supported by dispatching rule with four genes is proposed. Computational experiments are conducted based on multiple worker skills. The results showed that when the worker skills are the same, the genetic representation of the job name and part processing order is effective, whereas when the worker skills are different, the genetic representation of the cooperative process with the worker for each operation is effective.

Realization and Scheduling of Free Spot Assembly Method for Machine Tools Using Cooperative Industrial Robots

Kosuke Inoue,Hideki Aoyama 한국정밀공학회 2023 한국정밀공학회지 Vol.40 No.4

The manufacturing industry is increasingly demanding flexible manufacturing and existing manufacturing methods with fixed equipment do not meet this requirement. The free spot assembly system is an ultra-flexible method that responds to this demand, enabling spatiotemporal free assembly by conveying all necessary resources with automated guided vehicles (AGVs). Although some studies have proposed free spot assembly, free spot assembly feasibility for assembling heavy objects, such as machine tools, by aligning them at high precision has not been verified. Work hour shifts, differences in worker skill levels, and cooperative work with robots have also not been considered in free spot assembly scheduling. This paper presents elemental technologies for realizing a free spot assembly system, with a scheduling method where a genetic algorithm is supported by dispatching rules with six genes. The computational results reveal the effectiveness of the proposed algorithm.

Perioperative Evaluation of Respiratory Muscle Strength after Scoliosis Correction in Patients with Duchenne Muscular Dystrophy

Wataru Saito1,Kosuke Mizuno,Gen Inoue,Takayuki Imura1,Toshiyuki Nakazawa,Masayuki Miyagi,Eiki Shirasawa,Kentaro Uchida,Masashi Takaso 대한척추외과학회 2017 Asian Spine Journal Vol.11 No.5

Study Design: Retrospective cohort study. Purpose: To investigate the effect of spinal correction on respiratory muscle strength in patients with Duchenne muscular dystrophy (DMD). Overview of Literature: Several studies have reported that scoliosis correction in patients with DMD does not improve pulmonary function. In these studies, pulmonary function was evaluated using the traditional spirometric values of percent vital capacity (%VC) and percent forced vital capacity (%FVC). However, traditional spirometry may not be suitable for patients with DMD because the results can be influenced by patient fatigue or level of understanding. Therefore, we evaluated respiratory function focusing on respiratory muscle strength using maximal inspiratory pressure (MIP), maximal expiratory pressure (MEP), and sniff nasal inspiratory pressure (SNIP), in addition to %VC and %FVC. Methods: We retrospectively reviewed 16 patients with DMD who underwent spinal correction surgery between 2006 and 2011 at Kitasato University Hospital. All patients were males, and the mean age was 13.5 years. Respiratory muscle strength was evaluated using MIP, MEP, and SNIP. Measurements were obtained preoperatively and at 1 and 6 months postoperatively, and %VC and %FVC were obtained preoperatively and within 6 months postoperatively. Results: The mean preoperative and postoperative %VC values were 54.0% and 51.7%, whereas the mean %FVC values were 53.9% and 53.2%, respectively. The mean MIP, MEP, and SNIP values obtained preoperatively and at 1 and 6 months postoperatively were as follows: MIP, 40.5, 42.7 and 47.2 cm H2O; MEP, 26.0, 28.0, and 29.0 cm H2O; and SNIP, 33.4, 33.0, and 33.0 cm H2O; respectively. The mean MIP and MEP values significantly improved postoperatively. There were no significant differences in SNIP, %VC, or %FVC preand postoperatively. Conclusions: By focusing on respiratory muscle strength, our results suggest that scoliosis correction in patients with DMD might have a favorable effect on respiratory function.

Studies on Therapeutic Effects and Pathological Features of an Antithrombin Preparation in Septic Disseminated Intravascular Coagulation Patients

Yuichiro Sakamoto,Satoshi Inoue,Takashi Iwamura,Tomoko Yamashita,Atsushi Nakashima,Yoichi Nishimura,Hiroyuki Koami,Hisashi Imahase,Akiko Goto,Kosuke Chris Yamada,Kunihiro Mashiko,Hiroyuki Yokota 연세대학교의과대학 2013 Yonsei medical journal Vol.54 No.3

Purpose: Few reports have been made on the therapeutic effects as well as pathological features of an antithrombin preparation in patients diagnosed with septic disseminated intravascular coagulation (DIC) by the diagnostic criteria for acute DIC. Materials and Methods: A total of 88 sepsis patients who had received inpatient hospital care during the period from January 2000 through December 2008 were divided into two groups, an antithrombin group and a non-antithrombin group, to study the outcomes. Furthermore, the relationship between sepsis-related factors and DIC in 44 patients was studied. Results: The antithrombin group contained 34 patients, and the non-antithrombin group contained 54 patients. The outcomes were significantly better in the antithrombin group. The levels of protein C were low in DIC patients. Conclusion: Our results suggest that early administration of antithrombin might improve outcomes of septic DIC patients in the diagnostic criteria for Japanese Association for Acute Medicine acute DIC.

Usefulness of the Endotoxin Activity Assay to Evaluate the Degree of Lung Injury

Yuichiro Sakamoto,Satoshi Inoue,Takashi Iwamura,Tomoko Yamashita,Atsushi Nakashima,Hiroyuki Koami,Toru Miike,Mayuko Yahata,Hisashi Imahase,Akiko Goto,Showgo Narumi,Miho Ohta,Chris-Kosuke Yamada 연세대학교의과대학 2014 Yonsei medical journal Vol.55 No.4

Purpose: It has been reported that the Pulse Contour Cardiac Output (PiCCO) is very useful mainly in the field of intensive care and treatment to grasp the pathophysiologicalconditions of pulmonary edema because of its capability of obtainingdata such as Pulmonary Vascular Permeability Index (PVPI) and Extra VascularLung Water (EVLW). Furthermore, a high degree of usability of various markers has been reported for better understanding of the pathological conditions in cases with septicemia. Materials and Methods: The correlation between the cardiorespiratory status based upon the PiCCO monitor (EVLW and PVPI) and inflammatorymarkers including C reactive protein, procalcitonin (PC), and EndotoxinActivity Assay (EAA) were evaluated in 11 severe cases that required treatmentwith a respirator in an intensive care unit. Results: The EAA values were significantly higher in patients with abnormal EVLW at 0.46±0.20 compared to the normal EVLW group at 0.21±0.19 (p=0.0064). In a similar fashion, patients with abnormal PVPI values tended to have higher PC levels at 18.9±21.8 comparedto normal PVPI cases at 2.4±2.2 (p=0.0676). On the other hand, PVPI was significantly higher in the abnormal EAA group at 3.55±0.48 in comparison with the normal EAA group at 1.99±0.68 (p=0.0029). The abnormal EAA group tended to have higher PVPI values than the normal EAA group. Conclusion: The EAA is a measurement method designed to estimate the activity of endotoxins in the whole blood. Our results suggest that the EAA value, which had the greatest correlationwith lung disorders diagnosed by the PiCCO monitoring, reflects inflammatoryreactions predominantly in the lungs.