The advantages of scanning tunneling microscopy for the determination of the composition of the submonolayer metal/silicon interfaces has been demonstrated using Si(100)4×3-In and Si(111)4×1-In as sample reconstructions. It has been found that the Si(100)4×3-In unit cell is built of 7 In atoms and 6 Si atoms, while the Si(111)4×1-In unit cell contains 3 In atoms and 2 Si atoms in addition to the top Si(111) bilayer. The obtained quantitative information provides the ground for discussion of the plausible atomic arrangement of these reconstructions.

Using scanning tunneling microscopy (STM), we have studied the formation and structure of the $\mathrm{Si}\mathrm{}\left(100\right)2\times 3-\mathrm{Na}$ surface phase. It has been found that both the formation process and atomic arrangement of the $\mathrm{Si}\mathrm{}\left(100\right)2\times 3-\mathrm{Na}$ surface differ substantially from those known for $2\times 3$ reconstructions induced by other alkali and alkaline earth metals, since the Si(100)2×3-Na phase is the only one in which the top Si layer undergoes a substantial reordering. The Si-atom density in the $\mathrm{Si}\mathrm{}\left(100\right)2\times 3-\mathrm{Na}$ surface phase has been determined to be 1/3 ML. A structural model of the Si(100)2×3-Na reconstruction has been proposed, and its registry to STM images has been discussed.

One of the recent news headlines is that a pedestrian was killed by an autonomous vehicle because safety features in this vehicle did not detect an object on a road correctly. Due to this accident, some global automobile companies announced plans to postpone development of an autonomous vehicle. Furthermore, there is no doubt about the importance of safety features for autonomous vehicles. For this reason, our research goal is the development of a very safe and lightweight camera-based blind spot detection system, which can be applied to future autonomous vehicles. The blind spot detection system was implemented in open source software. Approximately 2000 vehicle images and 9000 non-vehicle images were adopted for training the Fully Connected Network (FCN) model. Other data processing concepts such as the Histogram of Oriented Gradients (HOG), heat map, and thresholding were also employed. We achieved 99.43% training accuracy and 98.99% testing accuracy of the FCN model, respectively. Source codes with respect to all the methodologies were then deployed to an off-the-shelf embedded board for actual testing on a road. Actual testing was conducted with consideration of various factors, and we confirmed 93.75% average detection accuracy with three false positives.