Films consisting of a silicon quantum dot superlattice were fabricated by alternating deposition of silicon rich silicon nitride and Si3N4 layers using an rf magnetron co-sputtering system. In order to use the silicon quantum dot super lattice structure for third generation multi junction solar cell applications, it is important to control the dot size. Moreover, silicon quantum dots have to be in a regularly spaced array in the dielectric matrix material for in order to allow for effective carrier transport. In this study, therefore, we fabricated silicon quantum dot superlattice films under various conditions and investigated crystallization behavior of the silicon quantum dot super lattice structure. Fourier transform infrared spectroscopy (FTIR) spectra showed an increased intensity of the 840 cm-1 peak with increasing annealing temperature due to the increase in the number of Si-N bonds. A more conspicuous characteristic of this process is the increased intensity of the 1100 cm-1 peak. This peak was attributed to annealing induced reordering in the films that led to increased Si-N4 bonding. X-ray photoelectron spectroscopy (XPS) analysis showed that peak position was shifted to higher bonding energy as silicon 2p bonding energy changed. This transition is related to the formation of silicon quantum dots. Transmission electron microscopy (TEM) and electron spin resonance (ESR) analysis also confirmed the formation of silicon quantum dots. This study revealed that post annealing at 1100˚C for at least one hour is necessary to precipitate the silicon quantum dots in the SiNx matrix.

Solar cells have been more intensely studied as part of the effort to find alternatives to fossil fuels as power sources.The progression of the first two generations of solar cells has seen a sacrifice of higher efficiency for more economic use ofmaterials. The use of a single junction makes both these types of cells lose power in two major ways: by the non-absorptionof incident light of energy below the band gap; and by the dissipation by heat loss of light energy in excess of the band gap.Therefore, multi junction solar cells have been proposed as a solution to this problem. However, the 1st and 2nd generation solarcells have efficiency limits because a photon makes just one electron-hole pair. Fabrication of all-silicon tandem cells using anSi quantum dot superlattice structure (QD SLS) is one possible suggestion. In this study, an SiOx matrix system was investigatedand analyzed for potential use as an all-silicon multi-junction solar cell. Si quantum dots with a super lattice structure (Si QDSLS) were prepared by alternating deposition of Si rich oxide (SRO; SiOx (x=0.8, 1.12)) and SiO2 layers using RF magnetronco-sputtering and subsequent annealing at temperatures between 800 and 1,100oC under nitrogen ambient. Annealing temperaturesand times affected the formation of Si QDs in the SRO film. Fourier transform infrared spectroscopy (FTIR) spectra and x-rayphotoelectron spectroscopy (XPS) revealed that nanocrystalline Si QDs started to precipitate after annealing at 1,100oC for onehour. Transmission electron microscopy (TEM) images clearly showed SRO/SiO2 SLS and Si QDs formation in each 4, 6, and8nm SRO layer after annealing at 1,100oC for two hours. The systematic investigation of precipitation behavior of Si QDsin SiO2 matrices is presented.

In this paper, we present a learning platform for robotic grasping in real world, in which actor-critic deep reinforcement learning is employed to directly learn the grasping skill from raw image pixels and rarely observed rewards. This is a challenging task because existing algorithms based on deep reinforcement learning require an extensive number of training data or massive computational cost so that they cannot be affordable in real world settings. To address this problems, the proposed learning platform basically consists of two training phases; a learning phase in simulator and subsequent learning in real world. Here, main processing blocks in the platform are extraction of latent vector based on state representation learning and disentanglement of a raw image, generation of adapted synthetic image using generative adversarial networks, and object detection and arm segmentation for the disentanglement. We demonstrate the effectiveness of this approach in a real environment.

A seismic performance assessment based on risk assessment is proposed considering risk factors and scenario, and thus the technique provides more practical and reasonable seismic performance assessment and reinforcement for the existing buildings more than ever before.

Currently the seismic performance evaluation for existing buildings are being made simply through the review of the analytical technique corresponding to design specifications on beam, column, shear wall, joint and so on. Also the seismic performance is evaluated to 4 steps as immediate occupancy level(IO), life safety level(LS), collapse prevention level(CP), collapse occurrence level(CO). However, in case of a structure being used for a number of years after construction, which may have partial damages and durability degradations may show significant differences between the analysis results and the actual behavior during an earthquake. Therefore, a improved seismic performance assessment is proposed based on risk assessment and seismic performance techniques since more systematic evaluation and condition assessment are required, and thus the practical and economical seismic retrofit measures for the existing facilities and for the reasonable seismic design as well are to be developed more than ever before.

One of the main problems of topological localization in a real indoor environment is variations in the environment caused by dynamic objects and changes in illumination. Another problem arises from the sense of topological localization itself. Thus, a robot must be able to recognize observations at slightly different positions and angles within a certain topological location as identical in terms of topological localization. In this paper, a possible solution to these problems is addressed in the domain of global topological localization for mobile robots, in which environments are represented by their visual appearance. Our approach is formulated on the basis of a probabilistic model called the Bayes filter. Here, marginalization of dynamics in the environment, marginalization of viewpoint changes in a topological location, and fusion of multiple visual features are employed to measure observations reliably, and action-based view transition model and action-associated topological map are used to predict the next state. We performed experiments to demonstrate the validity of our proposed approach among several standard approaches in the field of topological localization. The results clearly demonstrated the value of our approach.

Hierarchical Planning based on Abstraction of World Elements and Operators(HiPAWO) is proposed for mobile robots task planning, where abstraction of world elements is used for hierarchical planning and abstraction of operators is used for hierarchical decomposition of abstracted actions. Especially, a hierarchical domain theory based on JAH(Joint of Action Hierarchy)-graph is proposed to improve efficiency of planning, where a number of same actions are included in both adjacent hierarchical levels of domain theories to provide relationships between adjacent hierarchical levels. To show the validities of our proposed HiPAWO, experimental results are illustrated and will be compared with two other classical planning methods.