Machine learning has made breakthroughs in areas such as computer vision and natural language processing. In recent years, more and more research has been done on the application of machine learning on robotic grasping. This article summarizes the research progress of machine learning on robotic grasping, from the aspects of object grasping datasets, two main categories of methods that differ from the criteria for successful grasping with deep learning or reinforcement learning algorithm, discusses what current researches have done and the problems that have not yet been solved, and hopes to inspire new ideas in research of robotic grasping based on machine learning.
We present a new single-chip diaphragm-type Fabry-Perot microcavity pressure sensor with a novel single deeply corrugated diaphragm. Both analytical and experimental results have shown that some common issues, such as signal-averaging effect and cross-sensitivity to temperature with diaphragm-type Fabry-Perot microcavity pressure sensors, can be substantially alleviated by the proposed technique.
In this study, an analytical model, taking into account the coupled photoelastic and thermal-optical effects, is established to evaluate the temperature dependence of a single-chip silicon micromachined Fabry-Perot pressure sensor. The results show that temperature variation has significant impact on the micromachined Fabry-Perot pressure sensor with conventional flat diaphragm. A new membrane-type silicon micromachined Fabry-Perot pressure sensor with a novel deeply corrugated diaphragm is then proposed. The sensor is fabricated on a single-chip using both surface- and bulk-micromachining techniques. Both analytical and experimental results show that the cross-sensitivity to temperature of Fabry-Perot pressure sensors, can be substantially alleviated by the proposed single deeply corrugated diaphragm/mirror.
This paper presents results on optical cross-connect switches based on novel MEMS vertical mirrors. The switch consists of two mirror arrays to redirect optical beams from an input fiber array to an output fiber array. Each mirror is actuated by two electrostatic comb drive actuators, and can be rotated bi-directionally and perpendicularly to the chip surface. Finite element model (FEM) and Gaussian beam optics have been used to simulate and optimize the optical cross-connect switch architecture. Results have shown that the switch is much less constrained by the scaling distance of light propagation as the port count grows. However, the coupling efficiency is sensitive to angular alignment for large port-counts; thus mechanism for ensuring precise angular control of the micro-mirror is crucial for the proposed MEMS optical cross-connect switches.