Multitask learning for health condition identification and remaining useful life prediction: deep convolutional neural network approach

Predicting remaining useful life (RUL) is crucial for system maintenance. Condition monitoring makes not only degradation data available for RUL estimation but also categorized health status data for health state identification. However, RUL prediction has been treated as an independent process in most cases even though potential relevance exists with health status detection process. In this paper, we propose a convolution neural network based multi-task learning method to reflect the relatedness of RUL estimation with health status detection process. The proposed method applied to the C-MAPSS dataset for aero-engine unit prognostics supported superior performances to existing baseline models.

     

Case-based context ontology construction using fuzzy set theory for personalized service in a smart home environment

To provide context-based personalized services utilizing smart appliances in a smart home environment, we propose a framework for PersonAlized Service disCovery Using FuZZY-based CBR and Context Ontology (PASCUZZY). Basically, the PASCUZZY framework is implemented on case-based context ontology. To generate and manage the case instances on the case-based context ontology, we adopt the fuzzy set theory to transpose numerical-type context data sensed from the surrounding environment. The context is transposed to linguistic-type context instances on the context ontology. In addition, to formalize and manage the context and services as multi-attributed data, the context ontology was developed reflecting the structure of cases borrowed from case-based reasoning. Furthermore, we propose adaptation methods to adjust the generic fuzzy membership functions depending on the inhabitants’ context. It is performed by modifying the values of the membership number and/or modifying the numbers of the linguistic terms that are based on the inhabitants’ context to affect the membership numbers. The adapted membership functions return the personalized degree of memberships depending on the specialized context of a specific fuzzy variable. Inevitably, the number of cases on the case-based context ontology will be increased from time to time. We apply Ward’s method not only to reduce the search effort via a hierarchical clustering on the case-based context ontology but also to find the most similar service as a solution to the new context. To verify the superiority of the PASCUZZY framework, we perform two kinds of evaluations. First, we evaluate the effectiveness of the adaptation of the fuzzy membership functions. Second, we verify the effectiveness of the application of a clustering method to the case instances of the case-based context ontology to identify the most similar service. Results of the experiment verified the effectiveness and superiority of the PASCUZZY framework.     

The determination of polymer interfacial tension by drop image processing: Comparison of theory and experiment for the pair,poly(dimethyl siloxane)/polybutndiene

The interfacial tensions of immiscible binary blends of poly(dimethyl siloxane) (PDMS) and polybutadiene (PBD) have been determined as a function of molecular weight and temperature. The technique employed for these measurements takes advantage of recent advances in the determination and analysis of pendant fluid drop profiles. The experimental data are compared to the predictions of square gradient theories and theories based on the diffusion equation approach as developed by Helfand and co-workers. Qualitative agreement is obtained with both types of theory when the Flory-Huggins interaction parameter is taken to be comprised of two terms: a temperature independent term of entropic origin; and an enthalpic term that is inversely dependent upon the temperature.     

Immobilizing chromium in stainless steel slags with MnO addition

This work investigates the stabilizing impact of MnO on the leaching behavior of hazardous Cr-containing CaO-SiO₂-Al₂O₃-Cr₂O₃-MnO stainless steel slags after equilibrating at various elevated temperatures and evaluates the potential immobilization of Cr into a MnCr₂O₄ spinel phase from the existing Cr₂O₃ phase. The MnCr₂O₄ spinel phase was found to be an excellent Cr-stabilizer in stainless steel slags, where the leaching tendency of potentially hazardous Cr-related ions decreased with higher MnO content and lower equilibration temperatures within the range of 0 to 15 mass pct. and 1500 to 1300°C, respectively. Thermodynamic calculations by conducting the phase stability diagram also showed that the MnCr₂O₄ spinel phase was relatively stable and the Ca₃Si₂O₇ (Ca_3-xMn_xSi₂O₇) phase was relatively unstable compared with other crystal phases in acid extractant with pH value of 3.2. Combined with the scanning electron microscopy and X-ray powder diffraction results along with the thermodynamic calculations, the leached Cr-related ions was predominantly originating from the unstable amorphous glass phase.     

Design and Implementation of a Ubiquitous Robotic Space

This paper describes a concerted effort to design and implement a robotic service framework. The proposed framework is comprised of three conceptual spaces: physical, semantic, and virtual spaces, collectively referred to as a ubiquitous robotic space. We implemented a prototype robotic security application in an office environment, which confirmed that the proposed framework is an efficient tool for developing a robotic service employing IT infrastructure, particularly for integrating heterogeneous technologies and robotic platforms.     

Implementation of mmWave long-range backhaul for UAV-BS

Uncrewed aerial vehicles (UAVs) have become a vital element in nonterrestrial networks, especially with respect to 5G communication systems and beyond. The use of UAVs in support of 4G/5G base station (uncrewed aerial vehicle base station [UAV-BS]) has proven to be a practical solution for extending cellular network services to areas where conventional infrastructures are unavailable. In this study, we introduce a UAV-BS system that utilizes a high-capacity wireless backhaul operating in millimeter-wave frequency bands. This system can achieve a maximum throughput of 1.3 Gbps while delivering data at a rate of 300 Mbps, even at distances of 10 km. We also present the details of our testbed implementation alongside the performance results obtained from field tests.     

Facile synthesis of CsPbBr3/PbSe composite clusters

In this work, CsPbBr₃ and PbSe nanocomposites were synthesized to protect perovskite material from self-enlargement during reaction. UV absorption and photoluminescence (PL) spectra indicate that the addition of Se into CsPbBr₃ quantum dots modified the electronic structure of CsPbBr₃, increasing the band gap from 2.38 to 2.48 eV as the Cs:Se ratio increased to 1:3. Thus, the emission color of CsPbBr₃ perovskite quantum dots was modified from green to blue by increasing the Se ratio in composites. According to X-ray diffraction patterns, the structure of CsPbBr₃ quantum dots changed from cubic to orthorhombic due to the introduction of PbSe at the surface. Transmission electron microscopy and X-ray photoemission spectroscopy confirmed that the atomic distribution in CsPbBr₃/PbSe composite clusters is uniform and the composite materials were well formed. The PL intensity of a CsPbBr₃/PbSe sample with a 1:1 Cs:Se ratio maintained 50% of its initial intensity after keeping the sample for 81 h in air, while the PL intensity of CsPbBr₃ reduced to 20% of its initial intensity. Therefore, it is considered that low amounts of Se could improve the stability of CsPbBr₃ quantum dots.     

Learning decision trees with taxonomy of propositionalized attributes

We consider the problem of exploiting a taxonomy of propositionalized attributes in order to learn compact and robust classifiers. We introduce propositionalized attribute taxonomy guided decision tree learner (PAT-DTL), an inductive learning algorithm that exploits a taxonomy of propositionalized attributes as prior knowledge to generate compact decision trees. Since taxonomies are unavailable in most domains, we also introduce propositionalized attribute taxonomy learner (PAT-Learner) that automatically constructs taxonomy from data. PAT-DTL uses top-down and bottom-up search to find a locally optimal cut that corresponds to the literals of decision rules from data and propositionalized attribute taxonomy. PAT-Learner propositionalizes attributes and hierarchically clusters the propositionalized attributes based on the distribution of class labels that co-occur with them to generate a taxonomy. Our experimental results on UCI repository data sets show that the proposed algorithms can generate a decision tree that is generally more compact than and is sometimes comparably accurate to those produced by standard decision tree learners.     

A sales agent for part manufacturers: VMSA

The sales activity of most parts manufacturing companies is based on orders of buyers. The process of promotion, receipt and selection of orders of the parts manufacturers is closely coupled with the load status of the production lines. On deciding whether to accept an order or not, as well as negotiating with buyers, sales persons need information such as load and schedule of production lines, and manufacturability of the order. Manufacturability analysis, process planning, and scheduling are therefore key features in developing an agent of sales activity for the parts manufacturing business. In this paper, an intelligent agent, a virtual manufacturing-based sales agent (VMSA) with multi-agent architecture is proposed to support the sales activity for the parts manufacturers in the Internet environment. A VMSA has an integrated architecture of agent and virtual manufacturing. The prototype of VMSA has been developed for a machine parts manufacturing company that has lathe machines, drilling machines, cutting machines, and milling machines.