本文へジャンプ

成果報告書詳細
管理番号20180000000152
タイトル*平成29年度中間年報 IoT推進のための横断技術開発プロジェクト 超高効率データ抽出機能を有する学習型スマートセンシングシステムの研究開発
公開日2018/6/2
報告書年度2017 - 2017
委託先名技術研究組合NMEMS技術研究機構
プロジェクト番号P16007
部署名IoT推進部
和文要約
英文要約1-(1) Development of learning-based smart concentrators for ultra-efficient data collection and extraction: We built a learning-type concentrator system with learning and analytics functions by combining a concentrator, multiple sensors, and energy harvesters. We fabricated a smart sensing front-end circuit module that can be mounted on a smart sensor node. We also modularized the computational functions to meet the requirements for analyzing the sensor data, and prototyped related software. By collecting data with three sensor nodes to verify this system, we found that raw data could be condensed by a factor of 30 as valuable information. 1-(2) Inspection of an algorithm for monitoring the status of industrial facilities: We installed various types of sensors in machines, such as air-conditioning pumps, in our own and a customer’s facilities. Based on the collected data, we constructed algorithms for monitoring the status of industrial facilities and algorithms for substituting current daily inspections depended on human senses. 1-(3) Highly reliable, low-power, wireless, smart multi-sensor nodes: We developed wireless smart sensor nodes able to operate the “smart sensing front end.” These new nodes enabled the dynamic control of sensing parameters. We also developed nodes equipped with infra-red array sensors. Lastly, we created wireless modules with a mesh network system(sub-GHz) by improving the capabilities of the network layer. 2-(1) Development of low-power gas sensors to detect equipment overheat: Three signature gases were selected for use for the proactive detection of equipment overheats in factories and plants. Gas chromatography was used to identify the candidate gas species and narrow down the candidates to the final three. The feasibility of detecting the gases was confirmed by observing the sensitivity profiles from multiple catalytic sensor heads. 2-(2) A low-power IR sensor array for thermal imaging: A new operation algorithm was applied to reconfigure the imaging pixels and temperature resolution of a low-power IR sensor array for thermal imaging. The average power consumption of the sensor array was estimated to be 200 µW or less, which suggested that the algorithm could be adapted to enhance the sampling rate by deliberately limiting the number of imaging pixels. 3 High-efficiency MEMS vibrational energy harvesters: A novel energy harvester mechanism was developed using a series of symmetrically charged comb-drive electrodes coated with solid-ion electrets made of potassium-doped thermal silicon oxide. A new design for suspensions was proposed to reduce the device footprint to a scale compatible with MEMS foundry processes. A vacuum packaging process was also developed to enhance mechano-electric energy conversion. The proposed technology enables high-efficiency conversion from the harmonic vibrations of a mechanical compressor operating at 0.15 G at around 100 Hz.
ダウンロード成果報告書データベース(ユーザ登録必須)から、ダウンロードしてください。

▲トップに戻る