Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560406
Michael Danielczuk, Jeffrey Mahler, Christopher Correa, Ken Goldberg
To facilitate automated bin picking when parts cannot be grasped, pushing actions have the potential to separate objects and move them away from bin walls and corners. In the context of the Dexterity Network (Dex-Net) robot grasping framework, we present two novel push policies based on targeting free space and diffusing clusters, and compare them to three earlier policies using four metrics. We evaluate these in simulation using Bullet Physics on a dataset of over 1,000 synthetic pushing scenarios. Pushing outcomes are evaluated by comparing the quality of the best available grasp action before and after each push using analytic grasp metrics. Experiments conducted on scenarios in which Dex-Net could not successfully grasp objects suggest that pushing can increase the probability of executing a successful grasp by more than 15%. Furthermore, in cases where grasp quality can be improved, the new policies outperform a quasi-random baseline by nearly 2 times. In physical experiments on an ABB YuMi, the highest performing push policy increases grasp quality by 24%.
{"title":"Linear Push Policies to Increase Grasp Access for Robot Bin Picking","authors":"Michael Danielczuk, Jeffrey Mahler, Christopher Correa, Ken Goldberg","doi":"10.1109/COASE.2018.8560406","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560406","url":null,"abstract":"To facilitate automated bin picking when parts cannot be grasped, pushing actions have the potential to separate objects and move them away from bin walls and corners. In the context of the Dexterity Network (Dex-Net) robot grasping framework, we present two novel push policies based on targeting free space and diffusing clusters, and compare them to three earlier policies using four metrics. We evaluate these in simulation using Bullet Physics on a dataset of over 1,000 synthetic pushing scenarios. Pushing outcomes are evaluated by comparing the quality of the best available grasp action before and after each push using analytic grasp metrics. Experiments conducted on scenarios in which Dex-Net could not successfully grasp objects suggest that pushing can increase the probability of executing a successful grasp by more than 15%. Furthermore, in cases where grasp quality can be improved, the new policies outperform a quasi-random baseline by nearly 2 times. In physical experiments on an ABB YuMi, the highest performing push policy increases grasp quality by 24%.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"135 1","pages":"1249-1256"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79581273","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560427
Reza Haghighi, M. Rasouli, Syeda Mariam Ahmed, K. P. Tan, A. Mamun, C. Chew
Identifying a workpiece in industrial processes using depth sensors has received increasing attention over the past few years. However, this is a challenging task particularly when the object is large or cluttered. In these scenarios, captured point clouds do not provide sufficient information to detect the object. To address this issue, we present a hierarchical fragment matching method for 3D object detection and pose estimation. We build a library of object fragments by scanning the object from different viewpoints. A descriptor, named Clustered Centerpoint Feature Histogram (CCFH), is proposed to compute the features for each fragment. The proposed method aims to enhance the robustness of the existing Clustered Viewpoint Feature Histogram (CVFH) descriptor. Subsequently, an Extreme Learning Machine (ELM) classifier is applied to identify the matched segments between the scene and the library of fragments. Finally, the pose of the object in the scene is estimated using the matched segments. Unlike existing approaches that require the CAD model of the object or pre-registration process, the proposed method directly use the scanned point clouds of the object. The experimental results are presented to illustrate the performance of the proposed method.
{"title":"Depth-based Object Detection using Hierarchical Fragment Matching Method","authors":"Reza Haghighi, M. Rasouli, Syeda Mariam Ahmed, K. P. Tan, A. Mamun, C. Chew","doi":"10.1109/COASE.2018.8560427","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560427","url":null,"abstract":"Identifying a workpiece in industrial processes using depth sensors has received increasing attention over the past few years. However, this is a challenging task particularly when the object is large or cluttered. In these scenarios, captured point clouds do not provide sufficient information to detect the object. To address this issue, we present a hierarchical fragment matching method for 3D object detection and pose estimation. We build a library of object fragments by scanning the object from different viewpoints. A descriptor, named Clustered Centerpoint Feature Histogram (CCFH), is proposed to compute the features for each fragment. The proposed method aims to enhance the robustness of the existing Clustered Viewpoint Feature Histogram (CVFH) descriptor. Subsequently, an Extreme Learning Machine (ELM) classifier is applied to identify the matched segments between the scene and the library of fragments. Finally, the pose of the object in the scene is estimated using the matched segments. Unlike existing approaches that require the CAD model of the object or pre-registration process, the proposed method directly use the scanned point clouds of the object. The experimental results are presented to illustrate the performance of the proposed method.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"57 1","pages":"780-785"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79807142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560340
Florian Müller, Jan Janetzky, Uwe Behrnd, J. Jäkel, Ulrike Thomas
In this paper a novel type of variable impedance control (VIC) is presented. The controller adjusts the impedance depending on the force input of the user. In this way it is easy to accelerate and decelerate. Additionally, for high velocity the damping decreases and vice versa. This approach could be interpreted as a combination of acceleration-dependent VIC and velocity-dependent VIC. To guarantee stability, a stability observer is introduced. The observer is based on a model which describes a combined passive and active behavior of the user. In addition, we present a user study with 45 participants where the differences between VIC, VIC with stability observer and a pure admittance controller were investigated. The results show an improvement of the VIC with stability observer in relation to the pure admittance controller among different categories. With both the variable impedance controller and the variable impedance controller with stability observer, the participants significantly improved their times in comparison to the pure admittance controller, while they maintained the same level of precision. Also the workload was considerably smaller and the user comfort increased with both controllers compared to the usage of the pure admittance controller.
{"title":"User Force-Dependent Variable Impedance Control in Human-Robot Interaction","authors":"Florian Müller, Jan Janetzky, Uwe Behrnd, J. Jäkel, Ulrike Thomas","doi":"10.1109/COASE.2018.8560340","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560340","url":null,"abstract":"In this paper a novel type of variable impedance control (VIC) is presented. The controller adjusts the impedance depending on the force input of the user. In this way it is easy to accelerate and decelerate. Additionally, for high velocity the damping decreases and vice versa. This approach could be interpreted as a combination of acceleration-dependent VIC and velocity-dependent VIC. To guarantee stability, a stability observer is introduced. The observer is based on a model which describes a combined passive and active behavior of the user. In addition, we present a user study with 45 participants where the differences between VIC, VIC with stability observer and a pure admittance controller were investigated. The results show an improvement of the VIC with stability observer in relation to the pure admittance controller among different categories. With both the variable impedance controller and the variable impedance controller with stability observer, the participants significantly improved their times in comparison to the pure admittance controller, while they maintained the same level of precision. Also the workload was considerably smaller and the user comfort increased with both controllers compared to the usage of the pure admittance controller.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"21 1","pages":"1328-1335"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84666577","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560582
Michael Jäntsch, Naresh N. Nandola, Li Wang, M. Hakenberg, Ulrich Münz
In this work, we present an efficient planning algorithm for flexible manufacturing industries. In particular, we modified a traditional branch and bound approach to be used in a receding horizon manner by adopting the terminal cost concept from model predictive control domain. Thus, the proposed algorithm combines best practices from traditional planning and scheduling as well as from process control. The efficacy of the proposed algorithm is demonstrated on job shop problems of different sizes. Results are compared with traditional branch and bound based planning. The initial results are encouraging and demonstrate superior performance as well as scalability for large problems.
{"title":"Enhanced branch and bound approach for receding horizon based planning","authors":"Michael Jäntsch, Naresh N. Nandola, Li Wang, M. Hakenberg, Ulrich Münz","doi":"10.1109/COASE.2018.8560582","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560582","url":null,"abstract":"In this work, we present an efficient planning algorithm for flexible manufacturing industries. In particular, we modified a traditional branch and bound approach to be used in a receding horizon manner by adopting the terminal cost concept from model predictive control domain. Thus, the proposed algorithm combines best practices from traditional planning and scheduling as well as from process control. The efficacy of the proposed algorithm is demonstrated on job shop problems of different sizes. Results are compared with traditional branch and bound based planning. The initial results are encouraging and demonstrate superior performance as well as scalability for large problems.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"4 1","pages":"160-163"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80199850","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560515
Zhengcai Cao, Lu Liu, Meng Zhou
Short-term load forecasting (STLF) plays a very important role in the power system scheduling of smart grid. In this paper, a variable weight combined load forecasting model is proposed, effectively improves the accuracy of short-term load forecasting. A prediction model is presented by combining there single prediction models, i.e. random forest, extreme learning machine and Elman neural network. Then a bird swarm-based intelligent algorithm is utilized to solve the weighting problem among them. Experimental results demonstrate that the new constructed prediction model has higher prediction accuracy than any single load forecasting model.
{"title":"A Combined Model for Short-term Load Forecasting Based on Bird Swarm Algorithm","authors":"Zhengcai Cao, Lu Liu, Meng Zhou","doi":"10.1109/COASE.2018.8560515","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560515","url":null,"abstract":"Short-term load forecasting (STLF) plays a very important role in the power system scheduling of smart grid. In this paper, a variable weight combined load forecasting model is proposed, effectively improves the accuracy of short-term load forecasting. A prediction model is presented by combining there single prediction models, i.e. random forest, extreme learning machine and Elman neural network. Then a bird swarm-based intelligent algorithm is utilized to solve the weighting problem among them. Experimental results demonstrate that the new constructed prediction model has higher prediction accuracy than any single load forecasting model.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"1 1","pages":"791-796"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77771005","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560494
Ziyou Zhang, Qianchuan Zhao, Wen Yang
In intelligent buildings, faults occurring to sensors are common due to external interference or the limited service life of sensors. In this paper, we propose a distributed fault detection algorithm based on energy balance. With the increase of building scale, the algorithm has higher efficiency and better adaptability than the commonly used centralized fault detection methods. The algorithm is suitable for failures of disconnected sensors or devices, and makes judgements on whether the devices and sensors are faulty by the energy balance relations of adjacent nodes. Theoretical analysis and simulation show that this algorithm has lower time complexity than centralized algorithm which is also based on energy balance.
{"title":"A Distributed Algorithm for Sensor Fault Detection","authors":"Ziyou Zhang, Qianchuan Zhao, Wen Yang","doi":"10.1109/COASE.2018.8560494","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560494","url":null,"abstract":"In intelligent buildings, faults occurring to sensors are common due to external interference or the limited service life of sensors. In this paper, we propose a distributed fault detection algorithm based on energy balance. With the increase of building scale, the algorithm has higher efficiency and better adaptability than the commonly used centralized fault detection methods. The algorithm is suitable for failures of disconnected sensors or devices, and makes judgements on whether the devices and sensors are faulty by the energy balance relations of adjacent nodes. Theoretical analysis and simulation show that this algorithm has lower time complexity than centralized algorithm which is also based on energy balance.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"65 1","pages":"756-761"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81477337","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560373
J. Cheon, Jinwook Kim, Hongju Kim, Soonman Kwon, Youngkiu Choi
This paper presents a systematic design method of a controller using the discrete fuzzy PI control and the particle swarm optimization (PSO). Unlike a conventional PI controller, the discrete fuzzy PI controller has variable gains according to its input variables. Generally, it is complicated to tune the parameters of a fuzzy controller because there are too many parameters which are strongly coupled. In the discrete-time domain, the discrete fuzzy PI controller is a superset of the conventional PI controller. And the initial parameters of the fuzzy PI controller are selected by using the inclusion relationship. And, for the sake of simplicity, only four rules are used to construct a nonlinear fuzzy control surface. The tuning parameters of the discrete fuzzy PI controller are optimized by using the PSO. To verify the effectiveness of the controller designed by using the discrete fuzzy PI control and the PSO, we applied it to a wind turbine pitch controller. As a result, the proposed controller has variable gains, unlike the PI controller, and make the pitch controller operate in boarder operating regions.
{"title":"Systematic Design Method of Controller using Discrete Fuzzy PI Control and Particle Swarm Optimization","authors":"J. Cheon, Jinwook Kim, Hongju Kim, Soonman Kwon, Youngkiu Choi","doi":"10.1109/COASE.2018.8560373","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560373","url":null,"abstract":"This paper presents a systematic design method of a controller using the discrete fuzzy PI control and the particle swarm optimization (PSO). Unlike a conventional PI controller, the discrete fuzzy PI controller has variable gains according to its input variables. Generally, it is complicated to tune the parameters of a fuzzy controller because there are too many parameters which are strongly coupled. In the discrete-time domain, the discrete fuzzy PI controller is a superset of the conventional PI controller. And the initial parameters of the fuzzy PI controller are selected by using the inclusion relationship. And, for the sake of simplicity, only four rules are used to construct a nonlinear fuzzy control surface. The tuning parameters of the discrete fuzzy PI controller are optimized by using the PSO. To verify the effectiveness of the controller designed by using the discrete fuzzy PI control and the PSO, we applied it to a wind turbine pitch controller. As a result, the proposed controller has variable gains, unlike the PI controller, and make the pitch controller operate in boarder operating regions.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"136 1","pages":"1581-1586"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90739502","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560419
Lorenzo Abbatecola, M. P. Fanti, G. Pedroncelli, W. Ukovich
Modern logistics receives increasing attention for planning and scheduling operations of transport systems that have to be resource efficient and environmentally sustainable. A well known model for solving routing problems with time windows and vehicles is the Vehicle Routing Problem with Time Windows (VRPTW). This paper proposes a VRPTW algorithm based on cluster first, route second methods. In particular, first, by using a graph partitioning Integer Linear Programming problem, the algorithm generates a number of clusters equal to the number of available vehicles. Then, each vehicle solves a Travelling Salesman Problem with Time Windows to compute its route in the assigned cluster. Numerous benchmark problems featuring different sizes, random customer locations and time window distributions are solved and compared with the optimal solution. Moreover, a real case study shows the efficiency of the solution method..
{"title":"A New Cluster-Based Approach for the Vehicle Routing Problem with Time Windows","authors":"Lorenzo Abbatecola, M. P. Fanti, G. Pedroncelli, W. Ukovich","doi":"10.1109/COASE.2018.8560419","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560419","url":null,"abstract":"Modern logistics receives increasing attention for planning and scheduling operations of transport systems that have to be resource efficient and environmentally sustainable. A well known model for solving routing problems with time windows and vehicles is the Vehicle Routing Problem with Time Windows (VRPTW). This paper proposes a VRPTW algorithm based on cluster first, route second methods. In particular, first, by using a graph partitioning Integer Linear Programming problem, the algorithm generates a number of clusters equal to the number of available vehicles. Then, each vehicle solves a Travelling Salesman Problem with Time Windows to compute its route in the assigned cluster. Numerous benchmark problems featuring different sizes, random customer locations and time window distributions are solved and compared with the optimal solution. Moreover, a real case study shows the efficiency of the solution method..","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"39 1","pages":"744-749"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86697288","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560493
P. Sommer, Florian Schellroth, M. Fischer, Jan Schlechtendahl
Recent developments in production systems towards Industrial Internet of Things (Industrial loT) require a flexible communication and a distributed, loose coupled system architecture. Using the peer-to-peer communication model in combination with a message-oriented middleware (MOM) is a common approach for these new systems, but there is no information available about which MOM to use for manufacturing applications. Therefore, in this paper a comprehensive comparison of four well established MOM in terms of qualitative and quantitative criteria is presented, exceeding previous work. Furthermore, a methodology is presented to select the best fitting MOM for an industrial application.
{"title":"Message-oriented Middleware for Industrial Production Systems","authors":"P. Sommer, Florian Schellroth, M. Fischer, Jan Schlechtendahl","doi":"10.1109/COASE.2018.8560493","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560493","url":null,"abstract":"Recent developments in production systems towards Industrial Internet of Things (Industrial loT) require a flexible communication and a distributed, loose coupled system architecture. Using the peer-to-peer communication model in combination with a message-oriented middleware (MOM) is a common approach for these new systems, but there is no information available about which MOM to use for manufacturing applications. Therefore, in this paper a comprehensive comparison of four well established MOM in terms of qualitative and quantitative criteria is presented, exceeding previous work. Furthermore, a methodology is presented to select the best fitting MOM for an industrial application.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"8 1","pages":"1217-1223"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87949745","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560550
Wei-Chian Tan, I. Chen, D. Pantazis, Sinno Jialin Pan
This paper presents an end-to-end learning approach based on latest CNN architectures and transfer learning to perform vision-based analysis of engineering designs. The specific application considered here is the design of pipe networks on-board ships or offshore platforms. Having a piping design in the form of an image, a framework known as Piping Net (PipNet) is introduced to understand the design and interpret if it complies with applicable engineering regulations. Designs and corresponding labels (compliant or non-compliant) are fed into an existing trained CNN in the form of images for transfer learning, with the subsequently obtained fine-tuned network called PipNet. Based on Regulation 12, Annex I Regulations for the Prevention of Pollution by Oil, International Convention for the Prevention of Pollution from Ships (MARPOL) and Rules for Classification of Ships of Lloyd's Register, two datasets containing 3,234 piping designs in the form of images were used for performance evaluation. The developed system demonstrates outstanding performance on these two challenging datasets.
{"title":"Transfer Learning with PipNet: For Automated Visual Analysis of Piping Design","authors":"Wei-Chian Tan, I. Chen, D. Pantazis, Sinno Jialin Pan","doi":"10.1109/COASE.2018.8560550","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560550","url":null,"abstract":"This paper presents an end-to-end learning approach based on latest CNN architectures and transfer learning to perform vision-based analysis of engineering designs. The specific application considered here is the design of pipe networks on-board ships or offshore platforms. Having a piping design in the form of an image, a framework known as Piping Net (PipNet) is introduced to understand the design and interpret if it complies with applicable engineering regulations. Designs and corresponding labels (compliant or non-compliant) are fed into an existing trained CNN in the form of images for transfer learning, with the subsequently obtained fine-tuned network called PipNet. Based on Regulation 12, Annex I Regulations for the Prevention of Pollution by Oil, International Convention for the Prevention of Pollution from Ships (MARPOL) and Rules for Classification of Ships of Lloyd's Register, two datasets containing 3,234 piping designs in the form of images were used for performance evaluation. The developed system demonstrates outstanding performance on these two challenging datasets.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"92 1","pages":"1296-1301"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82691414","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：