Recently, manufacturing concepts, such as personalized production and distributed manufacturing, have attracted attention owing to the ongoing revolution in industrial technology. Connected micro smart factories factory-as-a-service system with these new paradigms Industrial Internet of Things (IIoT) are inefficient terms cost production. To solve problems, a digital twin, which uses representation process, same configuration elements, synchronized information, functional units, was designed...

In this paper, we present a two-stage neural quality estimation model that uses multilevel task learning for translation (QE) at the sentence, word, and phrase levels.Our approach is based on an end-to-end stacked named Predictor-Estimator, which has two stages consisting of word prediction QE model.To efficiently train model, stack propagation method applied, thereby enabling us to jointly learn in single mode.In addition, deploy with propagation, where training examples available all...

Recently, quality estimation has been attracting increasing interest from machine translation researchers, aiming at finding a good estimator for the “quality” of output. The common approach is to treat problem as supervised regression/classification task using quality-annotated noisy parallel corpus, called data , training data. However, available size remains small, due too-expensive cost creating such In addition, most conventional approaches rely on manually designed features model...

A URC (ubiquitous robotic companion) is a concept for network-based service robot. It allows the robot to extend its functions and services by utilizing external sensor networks remote computing servers. also provides robot's at any time place. The requires not only hardware infrastructure such as ubiquitous or high-performance servers but software which resides above infrastructure. In this paper, authors introduce CAMUS (context-aware middleware system) part of infrastructure, system...

Purpose In this paper, a three-dimensional (3D) printer-based manufacturing line and supporting system, which supports personalized/customized for individual businesses or start-up companies, was studied to evaluate the practicality of using additive personalization/mass customization. Design/methodology/approach First, factory-as-a-service (FaaS) provides factory as service customers, proposed designed manufacture various products within distributed environment. This system includes 3D...

For translation quality estimation at word and sentence levels, this paper presents a novel approach based on BERT that recently has achieved impressive results various natural language processing tasks. Our proposed model is re-purposed for the uses multi-task learning sentence-level task word-level subtasks (i.e., source word, target gap). Experimental Quality Estimation shared of WMT19 show our systems competitive provide significant improvements over baseline.

This paper describes the recurrent neural network based model for translation quality estimation. Recurrent estimation consists of two parts. The first part using bidirectional networks generates information about whether each word in is properly translated. second another predicts final translation. We apply this to sentence, and phrase level WMT16 Quality Estimation Shared Task. Our results achieve excellent performance especially sentence phraselevel QE.

Network and Internet technology opens up another domain for building future CAD/CAM environments. The environment will be global, network-centric, spatially distributed. We present an approach network-centric feature-based modeling in a distributed design environment. presented combines the current technique with computing communication supporting product collaborative activities over network. is implemented client/server architecture, which Web-enabled feature clients, neutral model server,...

Abstract Network and Internet technology open up another domain for building future CAD/CAM environments. The environment will be global, network-centric, spatially distributed. In this paper, we present Web-enabled feature-based modeling in a distributed design environment. presented approach combines the current technique with computing communication supporting product collaborative activities over network. is implemented client/server architecture, which feature clients, neutral model...

Network and Internet technology opens up another domain for building a future CAD/CAM environment. The environment will be global, network-centric, spatially distributed. In this paper, we present web-enabled approach to feature-based modeling in distributed collaborative design presented combines the current technique with computing communication supporting product activities over net work. is implemented client/server architecture, which Web-enabled feature clients, neutral model server,...

Pre-trained language models have brought significant improvements in performance a variety of natural processing tasks. Most existing performing state-of-the-art results shown their approaches the separate perspectives data processing, pre-training tasks, neural network modeling, or fine-tuning. In this paper, we demonstrate how affect individually, and that model performs best on specific question answering task when those are jointly considered models. particular, propose an extended task,...

Factory as a Service is mass personalization service platform. The system consists of three platforms; portal cloud system, manufacturing execution platform and operation including IoT based instruments called micro smart factory. factory include almost all the process instruments; instrument such 3D printer, post-processing instruments, inspection instrument, transfer packing instruments. works in multi agent on middleware. Each transfers fundamental information product part command. This...

This paper proposes a dual-layer user model to generate descriptive service recommendations for user-adaptive robots. The represents preferences as the associative memory in bottom-layer and association rules top-layer. learning inference processes two layers, bottom-up rule extraction process, are explained. proposed was applied coffee menu recommendation system, quantitative qualitative performances of system were evaluated by comparison with non-descriptive random methods.

This paper presents a novel Web-based service-oriented framework to support collaborative product commerce. The coordinates processes design and manufacture in loosely coupled environment using Web services. We have developed service model OWL. It provides facilities for companies discover services that other provide. Choreography of is achieved the process grammar. Usage grammar helps configure dynamically, facilitating collaboration among companies. framework, therefore, truly distributed...

URC (ubiquitous robotic companion) is a new concept for network-based service robot system. In this paper, we introduce CAMUS (context-aware middleware systems) which server framework to support context-aware services robots. We discuss two major components of CAMUS: main and agent manager. The communication also proposed the between robots CAMUS. expect that will be utilized develop ubiquitous computing applications network based robot, smart space personal assistant software

Quality estimation is an important task in machine translation that has attracted increased interest recent years. A key problem translation-quality the lack of a sufficient amount quality annotated training data. To address this shortcoming, Predictor-Estimator was proposed recently by introducing “word prediction” as additional pre-subtask predicts current target word with consideration surrounding source and contexts, resulting two-stage neural model composed predictor estimator ....