Pub Date : 2018-08-01DOI: 10.1109/COASE.2018.8560546
Kilian Telschig, Andreas Schonberger, Alexander Knapp
Container technologies such as Docker and Linux Containers (lxc) have become common tools in modern software engineering practice. They provide a dynamic and lightweight mechanism for software isolation and resource control, e.g. for continuous integration jobs or as app execution context. We adapt containers to industrial domains to offer enhanced reliability and legacy compatibility for distributed embedded applications. We describe a cross-domain real-time container architecture for dependable distributed embedded applications with criticality of timing requirements ranging from hard to non real-time. Through containers the proposed architecture isolates the software components from the system and from each other and only provides resources and inter-component communication explicitly demanded in each component's description. This enforces the interfaces and enables quality assurance and legacy compatibility. We provide a platform-independent model of the real-time container architecture but also describe a concrete lxc-based realization which conforms to this model.
{"title":"A Real-Time Container Architecture for Dependable Distributed Embedded Applications","authors":"Kilian Telschig, Andreas Schonberger, Alexander Knapp","doi":"10.1109/COASE.2018.8560546","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560546","url":null,"abstract":"Container technologies such as Docker and Linux Containers (lxc) have become common tools in modern software engineering practice. They provide a dynamic and lightweight mechanism for software isolation and resource control, e.g. for continuous integration jobs or as app execution context. We adapt containers to industrial domains to offer enhanced reliability and legacy compatibility for distributed embedded applications. We describe a cross-domain real-time container architecture for dependable distributed embedded applications with criticality of timing requirements ranging from hard to non real-time. Through containers the proposed architecture isolates the software components from the system and from each other and only provides resources and inter-component communication explicitly demanded in each component's description. This enforces the interfaces and enables quality assurance and legacy compatibility. We provide a platform-independent model of the real-time container architecture but also describe a concrete lxc-based realization which conforms to this model.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"100 1","pages":"1367-1374"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87005260","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.2018.8560477
Hongyue Sun, Giulia Pedrielli, Guanglei Zhao, Andrea Bragagnolo, Chi Zhou, R. Pan, Wenyao Xu
This paper extends the conventional single-stage additive manufacturing (AM) processes to multi-STage distRibutEd AM systems (STREAMs). In STREAM, a batch of material produced at the pre-processing stage is jointly consumed by distributed AM printers, and then the printed parts are collected for the centralized post-processing. Such systems are widely encountered in AM processes such as energy-AM, metal-AM and bio-AM. Modeling and managing such complex systems have been challenging. We propose a novel framework for “cyber-coordinated simulation” to manage the hierarchical information in STREAM. This is important because simulation can be used to infuse data into predictive analytics, thus providing guidance for the optimization and control of STREAM operations. The proposed framework is hierarchical in nature, where single stage, multi-stage and distributed productions are modeled through the integration of different simulators. We demonstrate the proposed framework with simulation data from freeze nano printing AM processes.
{"title":"Cyber-coordinated Simulation Models for Multi-stage Additive Manufacturing of Energy Products","authors":"Hongyue Sun, Giulia Pedrielli, Guanglei Zhao, Andrea Bragagnolo, Chi Zhou, R. Pan, Wenyao Xu","doi":"10.1109/COASE.2018.8560477","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560477","url":null,"abstract":"This paper extends the conventional single-stage additive manufacturing (AM) processes to multi-STage distRibutEd AM systems (STREAMs). In STREAM, a batch of material produced at the pre-processing stage is jointly consumed by distributed AM printers, and then the printed parts are collected for the centralized post-processing. Such systems are widely encountered in AM processes such as energy-AM, metal-AM and bio-AM. Modeling and managing such complex systems have been challenging. We propose a novel framework for “cyber-coordinated simulation” to manage the hierarchical information in STREAM. This is important because simulation can be used to infuse data into predictive analytics, thus providing guidance for the optimization and control of STREAM operations. The proposed framework is hierarchical in nature, where single stage, multi-stage and distributed productions are modeled through the integration of different simulators. We demonstrate the proposed framework with simulation data from freeze nano printing AM processes.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"73 1","pages":"893-898"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89897363","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.2018.8560366
Z. Fei, Shiqi Li, Q. Chang, Junfeng Wang, Yaqin Huang
In a manufacturing system, the idle status of machine consuming huge amounts of energy cannot bring any added value. How to reduce the energy waste of idle period through the real time control of machine status has become a challenging goal in an energy-efficient manufacturing environment. To address this problem, we propose a fuzzy Petri net based fuzzy reasoning approach to reduce the idle period by switching the on/off status of machines. The approach uses the real time data collected from the system, which include the level of upstream and downstream buffers, as well as the working status of the machine. The fuzzy rules are described by analyzing the decision intention according to the human knowledge. Simulation experiments show that this approach can effectively reduce the energy consumption with accepted throughput loss for a serial manufacturing system.
{"title":"Fuzzy Petri Net Based Intelligent Machine Operation of Energy Efficient Manufacturing System","authors":"Z. Fei, Shiqi Li, Q. Chang, Junfeng Wang, Yaqin Huang","doi":"10.1109/COASE.2018.8560366","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560366","url":null,"abstract":"In a manufacturing system, the idle status of machine consuming huge amounts of energy cannot bring any added value. How to reduce the energy waste of idle period through the real time control of machine status has become a challenging goal in an energy-efficient manufacturing environment. To address this problem, we propose a fuzzy Petri net based fuzzy reasoning approach to reduce the idle period by switching the on/off status of machines. The approach uses the real time data collected from the system, which include the level of upstream and downstream buffers, as well as the working status of the machine. The fuzzy rules are described by analyzing the decision intention according to the human knowledge. Simulation experiments show that this approach can effectively reduce the energy consumption with accepted throughput loss for a serial manufacturing system.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"116 6 Vyp 2. Neurology and psychiatry of elderly 1","pages":"1593-1598"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90242732","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.2018.8560458
Felix Spenrath, A. Pott
The fast determination of collision-free grasps is a key aspect in random bin picking. Heuristic search algorithms provide a feasible solution to this problem, using statistical data on the likelihood of finding a valid solution on elements with certain parameters. In this paper, we propose the use of several neural networks in such algorithms to accelerate the search while preserving the reliability. This is done by training the neural networks on the heuristic search trees of previous situations and using the output of these neural networks as part of the heuristic function. Finally, the effect of these neural networks is experimentally analyzed with sensor data from a working bin picking system with an industrial dual arm robot and it is shown that the calculation time in this setup is reduced by up to 45%.
{"title":"Using Neural Networks for Heuristic Grasp Planning in Random Bin Picking","authors":"Felix Spenrath, A. Pott","doi":"10.1109/COASE.2018.8560458","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560458","url":null,"abstract":"The fast determination of collision-free grasps is a key aspect in random bin picking. Heuristic search algorithms provide a feasible solution to this problem, using statistical data on the likelihood of finding a valid solution on elements with certain parameters. In this paper, we propose the use of several neural networks in such algorithms to accelerate the search while preserving the reliability. This is done by training the neural networks on the heuristic search trees of previous situations and using the output of these neural networks as part of the heuristic function. Finally, the effect of these neural networks is experimentally analyzed with sensor data from a working bin picking system with an industrial dual arm robot and it is shown that the calculation time in this setup is reduced by up to 45%.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"1 1","pages":"258-263"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85912116","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.2018.8560588
Suhyun Cha, A. Weigl, Mattias Ulbrich, Bernhard Beckert, B. Vogel‐Heuser
Automated production systems (aPS) operate for a long time with continuous and incremental changes. However, the models for aPS have not been maintained along with these system changes or, even, have not been properly generated. Even though the regression verification technique reduces the effort of applying formal verification on the automation system evolution, there still remains what should be provided in a formal form for the verification: delta, which is the difference of the two versions of the software. In this paper, we propose a method for generating a formal model from preexisting software in IEC 61131–3 Sequential Function Chart language. Based on this, the developer is able to achieve delta description by revising it to reflect the change request and this formal description of delta could facilitate verifying delta formally.
{"title":"Achieving delta description of the control software for an automated production system evolution","authors":"Suhyun Cha, A. Weigl, Mattias Ulbrich, Bernhard Beckert, B. Vogel‐Heuser","doi":"10.1109/COASE.2018.8560588","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560588","url":null,"abstract":"Automated production systems (aPS) operate for a long time with continuous and incremental changes. However, the models for aPS have not been maintained along with these system changes or, even, have not been properly generated. Even though the regression verification technique reduces the effort of applying formal verification on the automation system evolution, there still remains what should be provided in a formal form for the verification: delta, which is the difference of the two versions of the software. In this paper, we propose a method for generating a formal model from preexisting software in IEC 61131–3 Sequential Function Chart language. Based on this, the developer is able to achieve delta description by revising it to reflect the change request and this formal description of delta could facilitate verifying delta formally.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"47 1","pages":"1170-1176"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82859060","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.2018.8560523
Florian Pfitzer, Julien Provost, Carina Mieth, Wolfgang Liertz
Unpredictable incoming orders and the required nesting process highly complicate production planning and scheduling in sheet metal job shop environments and cause extremely high lead times as well as intermediate stocks. For this, numerous advanced planning and scheduling (APS) algorithms exist, aiming at creating a globally optimized production schedule. Due to the complexity of the multi-objective optimization and the large amount of unforeseen shop-floor events, effective and applicable solutions have not been presented so far. This work introduces an event-driven rescheduling concept based on lean principles leading to a high responsiveness of the production process to any kind of deviation. The achieved, significantly smaller buffer occupancies enable shorter lead times and improved delivery time estimations. Excellent performance results of the rescheduling concept are shown in different simulation experiments. The presented concept can easily be implemented in any kind of sheet metal job shop and its respective IT infrastructure.
{"title":"Event-Driven Production Rescheduling in Job Shop Environments","authors":"Florian Pfitzer, Julien Provost, Carina Mieth, Wolfgang Liertz","doi":"10.1109/COASE.2018.8560523","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560523","url":null,"abstract":"Unpredictable incoming orders and the required nesting process highly complicate production planning and scheduling in sheet metal job shop environments and cause extremely high lead times as well as intermediate stocks. For this, numerous advanced planning and scheduling (APS) algorithms exist, aiming at creating a globally optimized production schedule. Due to the complexity of the multi-objective optimization and the large amount of unforeseen shop-floor events, effective and applicable solutions have not been presented so far. This work introduces an event-driven rescheduling concept based on lean principles leading to a high responsiveness of the production process to any kind of deviation. The achieved, significantly smaller buffer occupancies enable shorter lead times and improved delivery time estimations. Excellent performance results of the rescheduling concept are shown in different simulation experiments. The presented concept can easily be implemented in any kind of sheet metal job shop and its respective IT infrastructure.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"83 1","pages":"939-944"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84030268","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.2018.8560359
S. Löw, D. Obradovic
Nonlinear Model Predictive Control (NMPC) is an aspiring control method for the implementation of advanced controller behavior. The present work shows the symbolic math implementation of a mechatronic system model containing aerodynamic nonlinearities modeled by Feedforward Neural Networks. Gradients for the optimization are obtained efficiently by exploiting the feedforward property of the Neural Networks and symbolic computation. Current research on the implementation of damage metrics into the cost function is stated briefly. In order to achieve real-time capability, the method Real-time Iteration is used.
{"title":"Real-time Implementation of Nonlinear Model Predictive Control for Mechatronic Systems Using a Hybrid Model","authors":"S. Löw, D. Obradovic","doi":"10.1109/COASE.2018.8560359","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560359","url":null,"abstract":"Nonlinear Model Predictive Control (NMPC) is an aspiring control method for the implementation of advanced controller behavior. The present work shows the symbolic math implementation of a mechatronic system model containing aerodynamic nonlinearities modeled by Feedforward Neural Networks. Gradients for the optimization are obtained efficiently by exploiting the feedforward property of the Neural Networks and symbolic computation. Current research on the implementation of damage metrics into the cost function is stated briefly. In order to achieve real-time capability, the method Real-time Iteration is used.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"18 1","pages":"164-167"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81968891","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.2018.8560496
Jason Li, Jonatan Berglund, Felix Auris, Atieh Hanna, J. Vallhagen, K. Åkesson
A digital twin of a production system consists of geometric, kinematic and logical models of the physical system. One of the key challenges is to keep the digital twin up-to-date with changes of the real one. Today, laser scanning is the de-facto standard used to keep the geometry of the digital model synchronized. In recent years, advancements in the performance of Graphic Processing Units (GPUs) and the availability of cheap high-resolution digital cameras have made photogrammetry a viable alternative to laser scanning for building digital 3D-models. In this study, we investigate how photogrammetry competes against laser-scanning by comparing their results in form of point-clouds.
{"title":"Evaluation of Photogrammetry for Use in Industrial Production Systems","authors":"Jason Li, Jonatan Berglund, Felix Auris, Atieh Hanna, J. Vallhagen, K. Åkesson","doi":"10.1109/COASE.2018.8560496","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560496","url":null,"abstract":"A digital twin of a production system consists of geometric, kinematic and logical models of the physical system. One of the key challenges is to keep the digital twin up-to-date with changes of the real one. Today, laser scanning is the de-facto standard used to keep the geometry of the digital model synchronized. In recent years, advancements in the performance of Graphic Processing Units (GPUs) and the availability of cheap high-resolution digital cameras have made photogrammetry a viable alternative to laser scanning for building digital 3D-models. In this study, we investigate how photogrammetry competes against laser-scanning by comparing their results in form of point-clouds.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"21 1","pages":"414-420"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78394665","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.2018.8560405
S. Scholz, A. Elkaseer, Tobias Muller, U. Gengenbach, V. Hagenmeyer
This paper presents an innovative approach to the development of a smart fully-digital manufacturing system based on a modular and reconfigurable production concept. In particular, multiple “plug and play” manufacturing modules, i.e. functional printing, laser processing and welding, in addition to a positioning control unit and quality inspection system, are exploited in an agile manufacturing platform combined with a knowledge-based framework, termed “3D-I”. This enables the production of tailored laminated parts, made up of stacks of functionalised layers of polymer films, with intricate 3D micro features. However, since no tool or mask making is needed, a medium to small lot-size and even one-off parts can be produced in a cost-effective manner. For the evaluation of the “3D-I” approach, a case study of micro-fluidic chips, exemplifying functional parts, are fabricated. The results prove the feasibility of the developed smart system to produce micro-devices with pre-defined specifications. In addition, the knowledge-based manufacturing system demonstrates its potential to offer profitable production scenarios of microdevices, with high flexibility and scalability, outside the area of mass production.
{"title":"Smart modular reconfigurable fully-digital manufacturing system with a knowledge-based framework: example of a fabrication of microfluidic chips","authors":"S. Scholz, A. Elkaseer, Tobias Muller, U. Gengenbach, V. Hagenmeyer","doi":"10.1109/COASE.2018.8560405","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560405","url":null,"abstract":"This paper presents an innovative approach to the development of a smart fully-digital manufacturing system based on a modular and reconfigurable production concept. In particular, multiple “plug and play” manufacturing modules, i.e. functional printing, laser processing and welding, in addition to a positioning control unit and quality inspection system, are exploited in an agile manufacturing platform combined with a knowledge-based framework, termed “3D-I”. This enables the production of tailored laminated parts, made up of stacks of functionalised layers of polymer films, with intricate 3D micro features. However, since no tool or mask making is needed, a medium to small lot-size and even one-off parts can be produced in a cost-effective manner. For the evaluation of the “3D-I” approach, a case study of micro-fluidic chips, exemplifying functional parts, are fabricated. The results prove the feasibility of the developed smart system to produce micro-devices with pre-defined specifications. In addition, the knowledge-based manufacturing system demonstrates its potential to offer profitable production scenarios of microdevices, with high flexibility and scalability, outside the area of mass production.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"67 1","pages":"1012-1017"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90768932","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.2018.8560699
Ishara Paranawithana, U-Xuan Tan, Liangjing Yang, Zhong Chen, K. Youcef-Toumi
This work proposes a fusion mechanism that overcomes the traditional limitations in vision-guided micromanipulation in plant cells. Despite the recent advancement in vision-guided micromanipulation, only a handful of research addressed the intrinsic issues related to micromanipulation in plant cells. Unlike single cell manipulation, the structural complexity of plant cells makes visual tracking extremely challenging. There is therefore a need to complement the visual tracking approach with trajectory data from the manipulator. Fusion of the two sources of data is done by combining the projected trajectory data to the image domain and template tracking data using a score-based weighted averaging approach. Similarity score reflecting the confidence of a particular localization result is used as the basis of the weighted average. As the projected trajectory data of the manipulator is not at all affected by the visual disturbances such as regional occlusion, fusing estimations from two sources leads to improved tracking performance. Experimental results suggest that fusion-based tracking mechanism maintains a mean error of 2.15 pixels whereas template tracking and projected trajectory data has a mean error of 2.49 and 2.61 pixels, respectively. Path B of the square trajectory demonstrated a significant improvement with a mean error of 1.11 pixels with 50% of the tracking ROI occluded by plant specimen. Under these conditions, both template tracking and projected trajectory data show similar performances with a mean error of 2.59 and 2.58 pixels, respectively. By addressing the limitations and unmet needs in the application of plant cell bio-manipulation, we hope to bridge the gap in the development of automatic vision-guided micromanipulation in plant cells.
{"title":"Scene-Adaptive Fusion of Visual and Motion Tracking for Vision-Guided Micromanipulation in Plant Cells","authors":"Ishara Paranawithana, U-Xuan Tan, Liangjing Yang, Zhong Chen, K. Youcef-Toumi","doi":"10.1109/COASE.2018.8560699","DOIUrl":"https://doi.org/10.1109/COASE.2018.8560699","url":null,"abstract":"This work proposes a fusion mechanism that overcomes the traditional limitations in vision-guided micromanipulation in plant cells. Despite the recent advancement in vision-guided micromanipulation, only a handful of research addressed the intrinsic issues related to micromanipulation in plant cells. Unlike single cell manipulation, the structural complexity of plant cells makes visual tracking extremely challenging. There is therefore a need to complement the visual tracking approach with trajectory data from the manipulator. Fusion of the two sources of data is done by combining the projected trajectory data to the image domain and template tracking data using a score-based weighted averaging approach. Similarity score reflecting the confidence of a particular localization result is used as the basis of the weighted average. As the projected trajectory data of the manipulator is not at all affected by the visual disturbances such as regional occlusion, fusing estimations from two sources leads to improved tracking performance. Experimental results suggest that fusion-based tracking mechanism maintains a mean error of 2.15 pixels whereas template tracking and projected trajectory data has a mean error of 2.49 and 2.61 pixels, respectively. Path B of the square trajectory demonstrated a significant improvement with a mean error of 1.11 pixels with 50% of the tracking ROI occluded by plant specimen. Under these conditions, both template tracking and projected trajectory data show similar performances with a mean error of 2.59 and 2.58 pixels, respectively. By addressing the limitations and unmet needs in the application of plant cell bio-manipulation, we hope to bridge the gap in the development of automatic vision-guided micromanipulation in plant cells.","PeriodicalId":6518,"journal":{"name":"2018 IEEE 14th International Conference on Automation Science and Engineering (CASE)","volume":"51 1","pages":"1434-1440"},"PeriodicalIF":0.0,"publicationDate":"2018-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79208392","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：