Pub Date : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817354
Ismail A Idowu, K. Nyarko, Otily Toutsop
A safety inspection is an on-site walk-through to identify potential hazards to occupants and personnel and options for remedial action. Although the Common approach for the safety inspection (Means of Egress MOE) is manual, this approach is ineffective and inexhaustive due to some inherent challenges: (1) infrequent inspection, and (2) inefficient use of trained human resources. To address these challenges, we introduced a Dual Temporal Buffer Differencing method. This computer vision-based approach automates the inspection of an interior building hallway (exit access) for an obstruction that may be a potential fire hazard. Our approach is important because it will mitigate the risk of a fire hazard to the building occupants by sensing and alerting the safety officer before a situation turns into an emergency. The performance of our proposed approach, the benefits, and the implementation challenges, were evaluated through a case study. The result demonstrates that our proposed Dual Temporal Buffer Differencing (DTBD) method can detect a potential fire hazard in the building exit access effectively and continuously. As a result, the approach can facilitate safety in the building and allow safety inspectors to plan more trained human resources.
{"title":"Computer Vision Method in Means of Egress Obstruction Detection","authors":"Ismail A Idowu, K. Nyarko, Otily Toutsop","doi":"10.1109/aiiot54504.2022.9817354","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817354","url":null,"abstract":"A safety inspection is an on-site walk-through to identify potential hazards to occupants and personnel and options for remedial action. Although the Common approach for the safety inspection (Means of Egress MOE) is manual, this approach is ineffective and inexhaustive due to some inherent challenges: (1) infrequent inspection, and (2) inefficient use of trained human resources. To address these challenges, we introduced a Dual Temporal Buffer Differencing method. This computer vision-based approach automates the inspection of an interior building hallway (exit access) for an obstruction that may be a potential fire hazard. Our approach is important because it will mitigate the risk of a fire hazard to the building occupants by sensing and alerting the safety officer before a situation turns into an emergency. The performance of our proposed approach, the benefits, and the implementation challenges, were evaluated through a case study. The result demonstrates that our proposed Dual Temporal Buffer Differencing (DTBD) method can detect a potential fire hazard in the building exit access effectively and continuously. As a result, the approach can facilitate safety in the building and allow safety inspectors to plan more trained human resources.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124362931","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817152
Fu-Shiung Hsieh
Ridesharing or shared mobility have been attracting significant attention in relevant research community. Most studies focus on how to match drivers and riders to minimize the overall travel distance based on their requirements. As cost savings is an essential function in ridesharing systems, allocation of cost savings has attracted researchers' attention recently. Several simple schemes have been proposed in the literature. For example, a simple scheme is to divide cost savings equally between driver and passengers in a ride. Another scheme is to allocate cost savings to participants proportional to their original travel distance. Although these simple schemes are easy to implement, there still lack a study that compare their effectiveness in ridesharing systems by applying different metaheuristic algorithms. In this paper, a hybrid meta-heuristic algorithm called hybrid Firefly-DE algorithm based on Differential Evolution and Firefly Algorithm will be adopted to match drivers and riders. We will compare three cost savings allocation schemes based on the numerical results. In our experiments, meta-heuristic algorithms are applied to find the matches to minimize the overall travel distance. The above schemes are then used to allocate cost savings among participants. The results indicate that the proportional cos savings allocation scheme is more effective than the other schemes to allocate cost savings equally between the drivers and the passengers.
{"title":"A Hybrid Firefly-DE algorithm for Ridesharing Systems with Cost Savings Allocation Schemes","authors":"Fu-Shiung Hsieh","doi":"10.1109/aiiot54504.2022.9817152","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817152","url":null,"abstract":"Ridesharing or shared mobility have been attracting significant attention in relevant research community. Most studies focus on how to match drivers and riders to minimize the overall travel distance based on their requirements. As cost savings is an essential function in ridesharing systems, allocation of cost savings has attracted researchers' attention recently. Several simple schemes have been proposed in the literature. For example, a simple scheme is to divide cost savings equally between driver and passengers in a ride. Another scheme is to allocate cost savings to participants proportional to their original travel distance. Although these simple schemes are easy to implement, there still lack a study that compare their effectiveness in ridesharing systems by applying different metaheuristic algorithms. In this paper, a hybrid meta-heuristic algorithm called hybrid Firefly-DE algorithm based on Differential Evolution and Firefly Algorithm will be adopted to match drivers and riders. We will compare three cost savings allocation schemes based on the numerical results. In our experiments, meta-heuristic algorithms are applied to find the matches to minimize the overall travel distance. The above schemes are then used to allocate cost savings among participants. The results indicate that the proportional cos savings allocation scheme is more effective than the other schemes to allocate cost savings equally between the drivers and the passengers.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114153979","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817299
Lap-Man Hoi, Yuqi Sun, S. Im
Sentence segmentation is important for improving the human readability of Automatic Speech Recognition (ASR) systems. Although it has been explored through numerous interdisciplinary studies, segmentation of Portuguese is still time-consuming due to the lack of efficient automatic segmentation methods and the reliance on qualified phonetic experts. This paper presents a novel algorithm that efficiently segments speech into sentences by learning the spectrogram of sentences through windows using a classification model developed with an Artificial Neural Network (ANN). Based on our experiments, the beginning part of a European Portuguese (EP) sentence enables better identification of the sentence's boundaries. In addition, a window frame of spectrogram constructed by the previous ending of 100 milliseconds (ms) and the subsequent beginning of 300 ms presents the best performance in the automatic sentence segmentation. As a result, the proposed algorithm can automatically segment Portuguese speech into sentences by analyzing its spectrogram without knowing the speech semantics.
{"title":"An Automatic Speech Segmentation Algorithm of Portuguese based on Spectrogram Windowing","authors":"Lap-Man Hoi, Yuqi Sun, S. Im","doi":"10.1109/aiiot54504.2022.9817299","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817299","url":null,"abstract":"Sentence segmentation is important for improving the human readability of Automatic Speech Recognition (ASR) systems. Although it has been explored through numerous interdisciplinary studies, segmentation of Portuguese is still time-consuming due to the lack of efficient automatic segmentation methods and the reliance on qualified phonetic experts. This paper presents a novel algorithm that efficiently segments speech into sentences by learning the spectrogram of sentences through windows using a classification model developed with an Artificial Neural Network (ANN). Based on our experiments, the beginning part of a European Portuguese (EP) sentence enables better identification of the sentence's boundaries. In addition, a window frame of spectrogram constructed by the previous ending of 100 milliseconds (ms) and the subsequent beginning of 300 ms presents the best performance in the automatic sentence segmentation. As a result, the proposed algorithm can automatically segment Portuguese speech into sentences by analyzing its spectrogram without knowing the speech semantics.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122113894","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817158
Shreehar Joshi, Eman Abdelfattah, Ryan Osgood
This work presents a classification problem to classify a movie's success based on features of a given movie. Two movies' datasets along with features generated from web scraping are utilized to generate the training and testing datasets. Four Machine Learning classifiers are applied to these datasets: Stochastic Gradient Descent, Random Forests, LinearSVC and Extra Trees. This study compares the performance metrics for these Machine Learning models on these two movies datasets and draws conclusions based on the results.
{"title":"Classification of Movie Success: A Comparison of Two Movie Datasets","authors":"Shreehar Joshi, Eman Abdelfattah, Ryan Osgood","doi":"10.1109/aiiot54504.2022.9817158","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817158","url":null,"abstract":"This work presents a classification problem to classify a movie's success based on features of a given movie. Two movies' datasets along with features generated from web scraping are utilized to generate the training and testing datasets. Four Machine Learning classifiers are applied to these datasets: Stochastic Gradient Descent, Random Forests, LinearSVC and Extra Trees. This study compares the performance metrics for these Machine Learning models on these two movies datasets and draws conclusions based on the results.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129588795","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817327
Benedicta Nana Esi Nyarko, Wu Bin, Jinzhi Zhou, George Kofi Agordzo, J. Odoom, Ebenezer Koukoyi
Since the outbreak of the coronavirus pandemic in December 2019, there has been increased interest in developing better facial recognition systems. This stems from the need to protect everyone from the spread of the virus. However, the measures taken to prevent the spread of the virus pose a challenge to security and surveillance systems as existing systems are unable to match faces with masks more efficiently. For this study, a custom dataset was generated due to the unavailability of a large face dataset for masked face recognition, and the existing datasets focused on Caucasians (white race faces) while Aethiopians (black race faces) were neglected. In this study, a comparative analysis was conducted between the AlexNet, ResNet-50, and Inception-V3 models to recognize faces with surgical masks, fabric masks, and N95 masks. The results of the study showed that the CNN models achieve excellent recognition accuracy for masked and unmasked faces. Analysis of the models' performance showed that the AlexNet model achieved 95.7%, ResNet-50 achieved 97.5%, and Inception-V3 also achieved 95.5%. From the study, ResNet-50 performed better than Inception-V3 and AlexNet models in recognizing masked faces.
{"title":"Comparative Analysis of AlexNet, Resnet-50, and Inception-V3 Models on Masked Face Recognition","authors":"Benedicta Nana Esi Nyarko, Wu Bin, Jinzhi Zhou, George Kofi Agordzo, J. Odoom, Ebenezer Koukoyi","doi":"10.1109/aiiot54504.2022.9817327","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817327","url":null,"abstract":"Since the outbreak of the coronavirus pandemic in December 2019, there has been increased interest in developing better facial recognition systems. This stems from the need to protect everyone from the spread of the virus. However, the measures taken to prevent the spread of the virus pose a challenge to security and surveillance systems as existing systems are unable to match faces with masks more efficiently. For this study, a custom dataset was generated due to the unavailability of a large face dataset for masked face recognition, and the existing datasets focused on Caucasians (white race faces) while Aethiopians (black race faces) were neglected. In this study, a comparative analysis was conducted between the AlexNet, ResNet-50, and Inception-V3 models to recognize faces with surgical masks, fabric masks, and N95 masks. The results of the study showed that the CNN models achieve excellent recognition accuracy for masked and unmasked faces. Analysis of the models' performance showed that the AlexNet model achieved 95.7%, ResNet-50 achieved 97.5%, and Inception-V3 also achieved 95.5%. From the study, ResNet-50 performed better than Inception-V3 and AlexNet models in recognizing masked faces.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128268127","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817263
Eric M Toth, M. Chowdhury
Cloud computing has become increasingly popular in the modern world. While it has brought many positives to the innovative technological era society lives in today, cloud computing has also shown it has some drawbacks. These drawbacks are present in the security aspect of the cloud and its many services. Security practices differ in the realm of cloud computing as the role of securing information systems is passed onto a third party. While this reduces managerial strain on those who enlist cloud computing it also brings risk to their data and the services they may provide. Cloud services have become a large target for those with malicious intent due to the high density of valuable data stored in one relative location. By soliciting help from the use of honeynets, cloud service providers can effectively improve their intrusion detection systems as well as allow for the opportunity to study attack vectors used by malicious actors to further improve security controls. Implementing honeynets into cloud-based networks is an investment in cloud security that will provide ever-increasing returns in the hardening of information systems against cyber threats.
{"title":"Honeynets and Cloud Security","authors":"Eric M Toth, M. Chowdhury","doi":"10.1109/aiiot54504.2022.9817263","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817263","url":null,"abstract":"Cloud computing has become increasingly popular in the modern world. While it has brought many positives to the innovative technological era society lives in today, cloud computing has also shown it has some drawbacks. These drawbacks are present in the security aspect of the cloud and its many services. Security practices differ in the realm of cloud computing as the role of securing information systems is passed onto a third party. While this reduces managerial strain on those who enlist cloud computing it also brings risk to their data and the services they may provide. Cloud services have become a large target for those with malicious intent due to the high density of valuable data stored in one relative location. By soliciting help from the use of honeynets, cloud service providers can effectively improve their intrusion detection systems as well as allow for the opportunity to study attack vectors used by malicious actors to further improve security controls. Implementing honeynets into cloud-based networks is an investment in cloud security that will provide ever-increasing returns in the hardening of information systems against cyber threats.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"60 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126868502","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817318
J. D. L. Cruz, Douglas Shimizu, K. George
Researchers are constantly striving to determine effective ways to detect and diagnose stress in patients as early as possible to prevent them from experiencing serious health consequences and complications. This study analyzed the subject's stress levels using EEG and fNIRS while they played a computer game that tested their ability to make accurate yet quick decisions. Trails were conducted to create a machine learning model to determine the varying levels of stress experienced by each subject. Blood oxygen levels, heart rate, and body temperature were also monitored and recorded. The EEG and fNIRS data was processed, tested, and verified using MATLAB to create the machine learning model. The data indicate that stress levels increased while the subject's quick decision-making skills were tested, and amplified as the difficulty of the computer game increased. The model accurately predicted and classified the level of stress an individual was under during each trial.
{"title":"EEG and fNIRS Analysis Using Machine Learning to Determine Stress Levels","authors":"J. D. L. Cruz, Douglas Shimizu, K. George","doi":"10.1109/aiiot54504.2022.9817318","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817318","url":null,"abstract":"Researchers are constantly striving to determine effective ways to detect and diagnose stress in patients as early as possible to prevent them from experiencing serious health consequences and complications. This study analyzed the subject's stress levels using EEG and fNIRS while they played a computer game that tested their ability to make accurate yet quick decisions. Trails were conducted to create a machine learning model to determine the varying levels of stress experienced by each subject. Blood oxygen levels, heart rate, and body temperature were also monitored and recorded. The EEG and fNIRS data was processed, tested, and verified using MATLAB to create the machine learning model. The data indicate that stress levels increased while the subject's quick decision-making skills were tested, and amplified as the difficulty of the computer game increased. The model accurately predicted and classified the level of stress an individual was under during each trial.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"58 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127886209","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817246
Maame Araba Vander-Pallen, P. Addai, Stuart Isteefanos, Tauheed Khan Mohd
Over time, many operating systems (OS) with a wide range of functions and features have emerged. As a consequence, they understand how each operating system has been built, which helps users' decisions while setting the operating system on their devices. As a result, a comparative research of various operating systems is required to offer specifics on the same as well as variation in fresh forms of OS to solve their problems. This paper explains the types of cyber attacks on the different types of operating systems. It analyses how operating systems become vulnerable and also how these vulnerabilities affect these operating systems. Our research highlights the impact that viruses have had on society since 2018, and we focus on the consequences that these viruses have. Our research has found a significant upward trend in the amount of cyber attacks in the last five years. We expect these numbers to continue their ascent in the future, especially in the cryptocurrency world.
{"title":"Survey on Types of Cyber Attacks on Operating System Vulnerabilities since 2018 onwards","authors":"Maame Araba Vander-Pallen, P. Addai, Stuart Isteefanos, Tauheed Khan Mohd","doi":"10.1109/aiiot54504.2022.9817246","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817246","url":null,"abstract":"Over time, many operating systems (OS) with a wide range of functions and features have emerged. As a consequence, they understand how each operating system has been built, which helps users' decisions while setting the operating system on their devices. As a result, a comparative research of various operating systems is required to offer specifics on the same as well as variation in fresh forms of OS to solve their problems. This paper explains the types of cyber attacks on the different types of operating systems. It analyses how operating systems become vulnerable and also how these vulnerabilities affect these operating systems. Our research highlights the impact that viruses have had on society since 2018, and we focus on the consequences that these viruses have. Our research has found a significant upward trend in the amount of cyber attacks in the last five years. We expect these numbers to continue their ascent in the future, especially in the cryptocurrency world.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132779369","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817236
M. Ghanbari, W. Kinsner, N. Sepehri
Electro-hydrostatic actuators (EHAs) are a type of hydraulic actuators which use pumps rather than valves to control the motion. As a result, they are more efficient than the valve-operated actuators. This paper presents an AI-based internal leakage detection algorithm for a single-rod EHA. Actuator internal leakage has been chosen to demonstrate the efficacy of the algorithm. Based on the sensitivity of various measures to varying levels of internal leakage, indicators are derived from the easy to obtain pressure measurements and a fault decision algorithm for quantifying the level of internal leakage in the actuator is established. This paper presents a new architecture of an artificial neural network (ANN) for detecting the existence of an internal leakage fault as labelled data. First, a sensitivity analysis is used to select a measure candidate for further research. Second, the measure chosen is analyzed using feature extraction methods. This step aims to extract hidden features to maximize the internal leakage fault detection. Finally, the fault detection algorithm classification efficiency is assessed by studying the detection rate of the proposed architecture. The experimental results show that the developed algorithm can detect internal leakage faults with 99.46% accuracy.
{"title":"Detection of Faults in Electro-Hydrostatic Actuators Using Feature Extraction Methods and an Artificial Neural Network","authors":"M. Ghanbari, W. Kinsner, N. Sepehri","doi":"10.1109/aiiot54504.2022.9817236","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817236","url":null,"abstract":"Electro-hydrostatic actuators (EHAs) are a type of hydraulic actuators which use pumps rather than valves to control the motion. As a result, they are more efficient than the valve-operated actuators. This paper presents an AI-based internal leakage detection algorithm for a single-rod EHA. Actuator internal leakage has been chosen to demonstrate the efficacy of the algorithm. Based on the sensitivity of various measures to varying levels of internal leakage, indicators are derived from the easy to obtain pressure measurements and a fault decision algorithm for quantifying the level of internal leakage in the actuator is established. This paper presents a new architecture of an artificial neural network (ANN) for detecting the existence of an internal leakage fault as labelled data. First, a sensitivity analysis is used to select a measure candidate for further research. Second, the measure chosen is analyzed using feature extraction methods. This step aims to extract hidden features to maximize the internal leakage fault detection. Finally, the fault detection algorithm classification efficiency is assessed by studying the detection rate of the proposed architecture. The experimental results show that the developed algorithm can detect internal leakage faults with 99.46% accuracy.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"130 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115469008","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 : 2022-06-06DOI: 10.1109/aiiot54504.2022.9817294
R. Dzhusupova, Richa Banotra, Jan Bosch, H. H. Olsson
Many organizations are looking for how to automate repetitive tasks to reduce manual work and free up resources for innovation. Machine Learning, especially Deep Learning, increases the chance of achieving this goal while working with technical documentation. Highly costly engineering hours can be saved, for example, by empowering the manual check with AI, which helps to reduce the total time for technical documents review. This paper proposes a way to substantially reduce the hours spent by process engineers reviewing P&IDs (Piping & Instrumentation Diagrams). The developed solution is based on a deep learning model for analyzing complex real-life engineering diagrams to find design errors - patterns that are combinations of high-level objects. Through the research on an extensive collection of P&ID files provided by McDermott, we prove that our model recognizes patterns representing engineering mistakes with high accuracy. We also describe our experience dealing with class-imbalance problems, labelling, and model architecture selection. The developed model is domain agnostic and can be re-trained on various schematic diagrams within engineering fields and, as well, could be used as an idea for other researchers to see whether similar solutions could be built for different industries.
{"title":"Pattern Recognition Method for Detecting Engineering Errors on Technical Drawings","authors":"R. Dzhusupova, Richa Banotra, Jan Bosch, H. H. Olsson","doi":"10.1109/aiiot54504.2022.9817294","DOIUrl":"https://doi.org/10.1109/aiiot54504.2022.9817294","url":null,"abstract":"Many organizations are looking for how to automate repetitive tasks to reduce manual work and free up resources for innovation. Machine Learning, especially Deep Learning, increases the chance of achieving this goal while working with technical documentation. Highly costly engineering hours can be saved, for example, by empowering the manual check with AI, which helps to reduce the total time for technical documents review. This paper proposes a way to substantially reduce the hours spent by process engineers reviewing P&IDs (Piping & Instrumentation Diagrams). The developed solution is based on a deep learning model for analyzing complex real-life engineering diagrams to find design errors - patterns that are combinations of high-level objects. Through the research on an extensive collection of P&ID files provided by McDermott, we prove that our model recognizes patterns representing engineering mistakes with high accuracy. We also describe our experience dealing with class-imbalance problems, labelling, and model architecture selection. The developed model is domain agnostic and can be re-trained on various schematic diagrams within engineering fields and, as well, could be used as an idea for other researchers to see whether similar solutions could be built for different industries.","PeriodicalId":409264,"journal":{"name":"2022 IEEE World AI IoT Congress (AIIoT)","volume":"110 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2022-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115718718","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：