Pub Date : 2024-02-22DOI: 10.14313/jamris/3-2023/18
Włodzimierz Kasprzak, Paweł Piwowarski
Research results on human activity classification in video are described, based on initial human skeleton estimation in video frames. Both single person actions and two-person interactions are considered. The initial skeleton data is estimated in selected video frames by OpenPose, HRNet or other dedicated library. Important contributions of presented work are computational steps of skeleton tracking and -refinement, and relational feature extraction from pairs of skeleton joints. It is shown, that this feature engineering significantly increases the classification accuracy. Regarding the final neural network encoder-classifier, two different architectures are designed and tested. The first solution is a lightweight MLP network, implementing the idea of a "mixture of pose experts". Several pose classifiers (experts) are trained independently on different time periods (snapshots) of single-person visual actions (or 2-person interactions), while the final classification is a time-related pooling of weighted expert classifications. All pose experts use the same deep encoding network. The second (middle weight) solution is based on a LSTM network.Both solutions are trained and tested on the action set of the well-known NTU RGB+D dataset, although only 2D data are used.Our results show comparable performance with some of the best reported STM- and CNN-based classifiers for this dataset. We conclude that by reducing the noise of skeleton data, highly successful lightweight- and midweight-approaches to visual activity recognition in image sequences can be achieved.
{"title":"Skeleton-based Human Activity Classification in Sparse Image Sequences","authors":"Włodzimierz Kasprzak, Paweł Piwowarski","doi":"10.14313/jamris/3-2023/18","DOIUrl":"https://doi.org/10.14313/jamris/3-2023/18","url":null,"abstract":"Research results on human activity classification in video are described, based on initial human skeleton estimation in video frames. Both single person actions and two-person interactions are considered. The initial skeleton data is estimated in selected video frames by OpenPose, HRNet or other dedicated library. Important contributions of presented work are computational steps of skeleton tracking and -refinement, and relational feature extraction from pairs of skeleton joints. It is shown, that this feature engineering significantly increases the classification accuracy. Regarding the final neural network encoder-classifier, two different architectures are designed and tested. The first solution is a lightweight MLP network, implementing the idea of a \"mixture of pose experts\". Several pose classifiers (experts) are trained independently on different time periods (snapshots) of single-person visual actions (or 2-person interactions), while the final classification is a time-related pooling of weighted expert classifications. All pose experts use the same deep encoding network. The second (middle weight) solution is based on a LSTM network.Both solutions are trained and tested on the action set of the well-known NTU RGB+D dataset, although only 2D data are used.Our results show comparable performance with some of the best reported STM- and CNN-based classifiers for this dataset. We conclude that by reducing the noise of skeleton data, highly successful lightweight- and midweight-approaches to visual activity recognition in image sequences can be achieved.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"5 8","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"139957844","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 : 2024-02-22DOI: 10.14313/jamris/3-2023/20
Bartłomiej Kulecki
In this paper, we propose a system for natural and intuitive interaction with the robot. Its purpose is to allow a person with no specialized knowledge and no training in robot programming to program a robotic arm.We utilize data from the RGB-D camera to segment the scene and detect objects. We also estimate the configuration of the operator's hand and the position of the visual marker to determine the intentions of the operator and the actions of the robot. To this end, we utilize trained neural networks and operations on the input point clouds. Also, voice commands are used to define or trigger the execution of the motion. Finally, we performed a set of experiments to show the properties of the proposed system.
{"title":"Multimodal Robot Programming Interface Based On RGB-D Perception and Neural Scene Understanding Modules","authors":"Bartłomiej Kulecki","doi":"10.14313/jamris/3-2023/20","DOIUrl":"https://doi.org/10.14313/jamris/3-2023/20","url":null,"abstract":"In this paper, we propose a system for natural and intuitive interaction with the robot. Its purpose is to allow a person with no specialized knowledge and no training in robot programming to program a robotic arm.We utilize data from the RGB-D camera to segment the scene and detect objects. We also estimate the configuration of the operator's hand and the position of the visual marker to determine the intentions of the operator and the actions of the robot. To this end, we utilize trained neural networks and operations on the input point clouds. Also, voice commands are used to define or trigger the execution of the motion. Finally, we performed a set of experiments to show the properties of the proposed system.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"12 s2","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-02-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140439175","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 : 2024-01-22DOI: 10.14313/jamris/2-2023/17
Megan Emmons, A. A. Maciejewski
Artificial swarms have the potential to provide robust, efficient solutions for a broad range of applications from assisting search and rescue operations to exploring remote planets. However, many fundamental obstacles still need to be overcome to bridge the gap between theory and application. In this characterization work, we demonstrate how a human rescuer can leverage minimal local observations of emergent swarm behavior to locate a lone survivor in maze-like environments. The simulated robots and rescuer have limited sensing and no communication capabilities to model a worst-case scenario. We then explore the impact of fundamental properties at the individual robot level on the utility of the emergent behavior to direct swarm design choices. We further demonstrate the relative robustness of the simulated robotic swarm by quantifying how reasonable probabilistic failure affects the rescue time in a complex environment. These results are compared to the theoretical performance of a single wall-following robot to further demonstrate the potential benefits of utilizing robotic swarms for rescue operations.
{"title":"Quantifying Swarm Resilience with Simulated Exploration of Maze-Like Environments","authors":"Megan Emmons, A. A. Maciejewski","doi":"10.14313/jamris/2-2023/17","DOIUrl":"https://doi.org/10.14313/jamris/2-2023/17","url":null,"abstract":"Artificial swarms have the potential to provide robust, efficient solutions for a broad range of applications from assisting search and rescue operations to exploring remote planets. However, many fundamental obstacles still need to be overcome to bridge the gap between theory and application. In this characterization work, we demonstrate how a human rescuer can leverage minimal local observations of emergent swarm behavior to locate a lone survivor in maze-like environments. The simulated robots and rescuer have limited sensing and no communication capabilities to model a worst-case scenario. We then explore the impact of fundamental properties at the individual robot level on the utility of the emergent behavior to direct swarm design choices. We further demonstrate the relative robustness of the simulated robotic swarm by quantifying how reasonable probabilistic failure affects the rescue time in a complex environment. These results are compared to the theoretical performance of a single wall-following robot to further demonstrate the potential benefits of utilizing robotic swarms for rescue operations.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"286 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140500007","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}
Forecasting solar radiation is one of the most useful impacts that can give us a deep vision on maintaining the integrity of solar systems. The availability and ease of use of the data make this process simpler. Predictions may be produced using various data sources. In fact, there are two different forms that can be identified. The first one was the use of historical solar radiation data, while the second one was the use of other meteorological parameters. The availability and choice of the data source can have an effect on the choice of the model and methods used. Our proposed article aims to take research as an example to review the solar radiation situation in Morocco and outline the methods of predicting solar radiation using different machine learning and deep learning methods like ANN, MLP, BPNN, DNN, and LSTM, which are used in different regions in Morocco.
{"title":"New Model of Photovoltaic System Adapted by a Digital MPPT Control and Radiation Predictions Using Deep Learning","authors":"A. Zouhri, M. el Mallahi","doi":"10.14313/2-2023/17","DOIUrl":"https://doi.org/10.14313/2-2023/17","url":null,"abstract":"Forecasting solar radiation is one of the most useful impacts that can give us a deep vision on maintaining the integrity of solar systems. The availability and ease of use of the data make this process simpler. Predictions may be produced using various data sources. In fact, there are two different forms that can be identified. The first one was the use of historical solar radiation data, while the second one was the use of other meteorological parameters. The availability and choice of the data source can have an effect on the choice of the model and methods used. Our proposed article aims to take research as an example to review the solar radiation situation in Morocco and outline the methods of predicting solar radiation using different machine learning and deep learning methods like ANN, MLP, BPNN, DNN, and LSTM, which are used in different regions in Morocco.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"377 2","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140501063","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 : 2024-01-22DOI: 10.14313//jamris/2-2023/15
Eduardo Garea-Llano
Alzheimer's disease is the most common form of dementia that can cause a brain neurological disorder with progressive memory loss as a result of brain cell damage. Prevention and treatment of disease is a key challenge in today's aging society. Accurate diagnosis of Alzheimer's disease plays an important role in patient management, especially in the early stages of the disease, because awareness of risk allows patients to undergo preventive measures even before brain damage occurs irreversible. Over the years, techniques such as statistical modeling or machine learning algorithms have been used to improve understanding of this condition. The objective of the work is the study of the methods of detection and progression of Alzheimer's disease through artificial intelligence techniques that have been proposed in the last three years.
{"title":"Update on the Study of Alzheimer´s Disease Through Artificial Intelligence Techniques","authors":"Eduardo Garea-Llano","doi":"10.14313//jamris/2-2023/15","DOIUrl":"https://doi.org/10.14313//jamris/2-2023/15","url":null,"abstract":"Alzheimer's disease is the most common form of dementia that can cause a brain neurological disorder with progressive memory loss as a result of brain cell damage. Prevention and treatment of disease is a key challenge in today's aging society. Accurate diagnosis of Alzheimer's disease plays an important role in patient management, especially in the early stages of the disease, because awareness of risk allows patients to undergo preventive measures even before brain damage occurs irreversible. \u0000Over the years, techniques such as statistical modeling or machine learning algorithms have been used to improve understanding of this condition. The objective of the work is the study of the methods of detection and progression of Alzheimer's disease through artificial intelligence techniques that have been proposed in the last three years.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"9 3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140499564","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 : 2024-01-22DOI: 10.14313/jamris/2-2023/11
Reynaldo Rosado, O. G. Toledano-López, Hector Gonzalez, A. J. Abreu, Yanio Hernandez
Recently, time series forecasting modelling in the Consumer Price Index (CPI) has attracted the attention of the scientific community. Several researches have tackled the problem of CPI prediction for their countries using statistical learning, machine learning and deep neural networks. The most popular approach to CPI in several countries is the Autoregressive Integrated Moving Average (ARIMA) due to the nature of the data. This paper addresses the Cuban CPI forecasting problem using Transformer with attention model over univariate dataset. The fine tuning of the lag parameter show that Cuban CPI have better performance with smalls lag and the best result was in $p=1$. Finally, the comparative results between ARIMA and our proposal show that the Transformer with attention has a very high performance despite having a small data set.
{"title":"Cuban Consumer Price Index Forecasting Trough Transformer with Attention","authors":"Reynaldo Rosado, O. G. Toledano-López, Hector Gonzalez, A. J. Abreu, Yanio Hernandez","doi":"10.14313/jamris/2-2023/11","DOIUrl":"https://doi.org/10.14313/jamris/2-2023/11","url":null,"abstract":"Recently, time series forecasting modelling in the Consumer Price Index (CPI) has attracted the attention of the scientific community. Several researches have tackled the problem of CPI prediction for their countries using statistical learning, machine learning and deep neural networks. The most popular approach to CPI in several countries is the Autoregressive Integrated Moving Average (ARIMA) due to the nature of the data. This paper addresses the Cuban CPI forecasting problem using Transformer with attention model over univariate dataset. The fine tuning of the lag parameter show that Cuban CPI have better performance with smalls lag and the best result was in $p=1$. Finally, the comparative results between ARIMA and our proposal show that the Transformer with attention has a very high performance despite having a small data set.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"136 2-3","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140500227","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 : 2024-01-22DOI: 10.14313/jamris/2-2023/14
Mamta Pandey, R. Litoriya, Prateek Pandey
During the past few years, the number of drones (unmanned aerial vehicles, or UAVs) manufactured and purchased has risen dramatically. It is predicted that it will continue to spread, making its use inevitable in all walks of life. Drone apps are therefore expected to overrun the app stores in the near future. The UAV's software is not being studied/researched despite several active research and studies being carried out in the UAV's hardware field. A large-scale empirical analysis of Google Play Store Platform apps connected to drones is being done in this direction. There are, however, a number of challenges with drone apps because of the lack of formal and specialised app development procedures. In this paper, eleven drone app issues have been identified. Then we applied the DEMATEL (Decision Making Trial and Evaluation Laboratory) method to analyse the drone app issues (DIs) and divide these issues into cause and effect groups. First, multiple experts assess the direct relationships between influential issues in drone apps. The evaluation results are presented in spherical fuzzy numbers (SFN). Secondly, convert the linguistic terms into SFN. Thirdly, based on DEMATEL, the cause-effect classifications of issues are obtained. Finally, the issues in the cause category are identified as DI’s in drone apps. The outcome of the research is compared with the other variants of DEMATEL like rough‐Z‐number‐based DEMATEL and spherical fuzzy number, and the comparative results suggest that Spherical Fuzzy-DEMATEL is the most fitting method to analyse the interrelationship of different issues in drone apps. The outcome of this work definitely assists the software industry in the successful identification of the critical issues where professionals and project managers could really focus.
在过去几年里，无人机（无人驾驶飞行器）的制造和购买数量急剧上升。据预测，无人机还将继续普及，各行各业都将不可避免地使用无人机。因此，预计在不久的将来，无人机应用程序将充斥应用程序商店。尽管在无人机硬件领域开展了多项积极的研究和调查，但无人机的软件尚未得到研究/调查。在这方面，正在对与无人机相关的 Google Play 商店平台应用程序进行大规模的实证分析。然而，由于缺乏正规和专业的应用程序开发流程，无人机应用程序面临着许多挑战。本文确定了 11 个无人机应用程序问题。然后，我们采用 DEMATEL（决策制定试验和评估实验室）方法对无人机应用程序问题（DIs）进行分析，并将这些问题分为因果组。首先，多位专家对无人机应用程序中具有影响力的问题之间的直接关系进行评估。评估结果用球形模糊数（SFN）表示。其次，将语言术语转换为 SFN。第三，基于 DEMATEL，得到问题的因果分类。最后，将原因类别中的问题确定为无人机应用程序中的 DI。研究成果与 DEMATEL 的其他变体（如基于粗糙 Z 数的 DEMATEL 和球形模糊数）进行了比较，比较结果表明，球形模糊 DEMATEL 是分析无人机应用程序中不同问题相互关系的最合适方法。这项工作的成果无疑有助于软件行业成功识别关键问题，使专业人员和项目经理能够真正关注这些问题。
{"title":"Application of the Spherical Fuzzy DEMATEL Model for Assessing the Drone Apps Issues","authors":"Mamta Pandey, R. Litoriya, Prateek Pandey","doi":"10.14313/jamris/2-2023/14","DOIUrl":"https://doi.org/10.14313/jamris/2-2023/14","url":null,"abstract":"During the past few years, the number of drones (unmanned aerial vehicles, or UAVs) manufactured and purchased has risen dramatically. It is predicted that it will continue to spread, making its use inevitable in all walks of life. Drone apps are therefore expected to overrun the app stores in the near future. The UAV's software is not being studied/researched despite several active research and studies being carried out in the UAV's hardware field. A large-scale empirical analysis of Google Play Store Platform apps connected to drones is being done in this direction. There are, however, a number of challenges with drone apps because of the lack of formal and specialised app development procedures. In this paper, eleven drone app issues have been identified. Then we applied the DEMATEL (Decision Making Trial and Evaluation Laboratory) method to analyse the drone app issues (DIs) and divide these issues into cause and effect groups. First, multiple experts assess the direct relationships between influential issues in drone apps. The evaluation results are presented in spherical fuzzy numbers (SFN). Secondly, convert the linguistic terms into SFN. Thirdly, based on DEMATEL, the cause-effect classifications of issues are obtained. Finally, the issues in the cause category are identified as DI’s in drone apps. The outcome of the research is compared with the other variants of DEMATEL like rough‐Z‐number‐based DEMATEL and spherical fuzzy number, and the comparative results suggest that Spherical Fuzzy-DEMATEL is the most fitting method to analyse the interrelationship of different issues in drone apps. The outcome of this work definitely assists the software industry in the successful identification of the critical issues where professionals and project managers could really focus.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"163 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140500216","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 : 2024-01-22DOI: 10.14313/jamris/2-2023/12
Maricruz Olazabal-Lugo, Luis F. Espinoza-Audelo, Ernesto León-Castro, L. A. Pérez-Arellano, Fabio Blanco-Mesa
This paper presents a new aggregation operator technique that uses the ordered weighted average (OWA), heavy aggregation operators, Hamming distance, and moving averages. This approach is called heavy ordered weighted moving average distance (HOWMAD). The main advantage of this operator is that it can use the characteristics of the HOWMA operator to under-or overestimate the results according to the expectations and the knowledge of the future scenarios, analyze the historical data of the moving average, and compare the different alternatives with the ideal results of the distance measures. Some of the main families and specific cases using generalized and quasi-arithmetic means are presented, such as the generalized heavy moving average distance and a generalized HOWMAD. This study develops an application of this operator in forecasting the sales growth rate for a commercial company. We find that it is possible to determine whether the company's objectives can be achieved or must be reevaluated in response to the actual situation and future expectations of the enterprise.
{"title":"Heavy Moving Average Distances in Sales Forecasting","authors":"Maricruz Olazabal-Lugo, Luis F. Espinoza-Audelo, Ernesto León-Castro, L. A. Pérez-Arellano, Fabio Blanco-Mesa","doi":"10.14313/jamris/2-2023/12","DOIUrl":"https://doi.org/10.14313/jamris/2-2023/12","url":null,"abstract":"This paper presents a new aggregation operator technique that uses the ordered weighted average (OWA), heavy aggregation operators, Hamming distance, and moving averages. This approach is called heavy ordered weighted moving average distance (HOWMAD). The main advantage of this operator is that it can use the characteristics of the HOWMA operator to under-or overestimate the results according to the expectations and the knowledge of the future scenarios, analyze the historical data of the moving average, and compare the different alternatives with the ideal results of the distance measures. Some of the main families and specific cases using generalized and quasi-arithmetic means are presented, such as the generalized heavy moving average distance and a generalized HOWMAD. This study develops an application of this operator in forecasting the sales growth rate for a commercial company. We find that it is possible to determine whether the company's objectives can be achieved or must be reevaluated in response to the actual situation and future expectations of the enterprise.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"298 4-5","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140499852","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 : 2024-01-22DOI: 10.14313/jamris/2-2023/16
Piotr Falkowski, C. Rzymkowski, Zbigniew Pilat
Contemporary international pandemic proved that a flexible approach towards work, trade and healthcare is not only favourable but a must. Hence, the devices enabling home-rehabilitation became one of the urgent needs of the medical market. The following overview is a part of an R&D project aimed at designing an exoskeleton and developing methods enabling effective home rehabilitation. It contains a comparison of current devices in terms of their kinematics, applications, weights, sizes, and integration with selected ICT technologies. The data is analysed regarding conclusions from qualitative research, based on in-depth interviews with physiotherapists and questionnaires organised beforehand. The investigation assesses whether commercial and developed devices enable feedback from a patient by all possible means; hence, if they could allow effective telerehabilitation. Moreover, their capabilities of increasing engagement and accelerating improvements by supervising techniques and measuring biomechanical parameters are evaluated. These outcomes are a base to set the constraints and requirements before designing an exoskeleton dedicated to home treatment.
{"title":"Analysis of Rehabilitation Systems in Regards to Requirements Towards Remote Home Rehabilitation Devices","authors":"Piotr Falkowski, C. Rzymkowski, Zbigniew Pilat","doi":"10.14313/jamris/2-2023/16","DOIUrl":"https://doi.org/10.14313/jamris/2-2023/16","url":null,"abstract":"Contemporary international pandemic proved that a flexible approach towards work, trade and healthcare is not only favourable but a must. Hence, the devices enabling home-rehabilitation became one of the urgent needs of the medical market. The following overview is a part of an R&D project aimed at designing an exoskeleton and developing methods enabling effective home rehabilitation. It contains a comparison of current devices in terms of their kinematics, applications, weights, sizes, and integration with selected ICT technologies. The data is analysed regarding conclusions from qualitative research, based on in-depth interviews with physiotherapists and questionnaires organised beforehand. The investigation assesses whether commercial and developed devices enable feedback from a patient by all possible means; hence, if they could allow effective telerehabilitation. Moreover, their capabilities of increasing engagement and accelerating improvements by supervising techniques and measuring biomechanical parameters are evaluated. These outcomes are a base to set the constraints and requirements before designing an exoskeleton dedicated to home treatment.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"142 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140500368","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 : 2024-01-22DOI: 10.14313/jamris/2-2023/13
M. Kamola
This paper presents a complete simulation and reinforcement learning solution to train mobile agents’ strategy of route tracking and avoiding mutual collisions. The aim was to achieve such functionality with limited resources, w.r.t. model input and model size itself. The designed models prove to keep agents safely on the track. Collision avoidance agent’s skills developed in the course of model training are primitive but rational. Small size of the model allows fast training with limited computational resources.
{"title":"A Compact DQN Model for Mobile Agents with Collision Avoidance","authors":"M. Kamola","doi":"10.14313/jamris/2-2023/13","DOIUrl":"https://doi.org/10.14313/jamris/2-2023/13","url":null,"abstract":"This paper presents a complete simulation and reinforcement learning solution to train mobile agents’ strategy of route tracking and avoiding mutual collisions. The aim was to achieve such functionality with limited resources, w.r.t. model input and model size itself. The designed models prove to keep agents safely on the track. Collision avoidance agent’s skills developed in the course of model training are primitive but rational. Small size of the model allows fast training with limited computational resources.","PeriodicalId":37910,"journal":{"name":"Journal of Automation, Mobile Robotics and Intelligent Systems","volume":"18 11","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"140499508","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：