現今有許多研究,透過蒐集感測器的訊號,建立機械元件的損耗模型,並利用該損耗模型,來判斷機械元件的壽命與使用狀況。但主要的問題是,如何將賣到世界各地的工具機,有效的蒐集工具機運作的物理訊號,並將物理訊號上傳至同一個雲端資料儲存中心,以進行後面的磨耗分析與損耗模型建立呢? 基於上述原因,本研究提出M2M架構的機械監測系統,並改良先前的研究,導入雲端資料平台,再將機械監測系統自主化。該機械監測系統包含多數之無線感測節點與伺服器。無線感測節點會蒐集工具機運作的物理訊號,透過無線傳輸的方式,將物理訊號傳送給伺服器。伺服器再使用行動通訊網路(3G或LTE),將工具機的物理訊號,上傳至Google的雲端資料儲存中心(Google App Engine, GAE)。本研究有在中正大學的機械工廠與實驗室,測試機械監測系統的穩定性。
Newly developed techniques for intelligent sensor systems make it possible to register the mechanical wear-out of parts, such as band saws, ball screws and gearbox reducers, by collecting working signals from them, such as vibrations and preload pressure and temperature changes. To build an accurate wear model, we need to log as many real signals as possible from numerous parts in machine tools. This raises a substantial problem: How can we collect a large number of real signals from the parts installed in many machine tools—which could be located anywhere in the world—and aggregate data to use in constructing a wearing model, as well as enabling remote systems analysis and send warnings if the parts are worn? In this study, based on our previous work, we design a M2M system to realize a cloud-based service that logs mechanical wear-out of parts. The proposed system is characterized by using a mobile network, such as GPRS /3G, to upload collected data without the need to pre-deploy conventional network equipment in factories; and taking advantages of off-the-shelf cloud platform services (Google Apps Engine, GAE) rather than maintaining a private database for logging the data, which reduces labor and equipment costs for the database. We design and implement this system and install it in a factory to evaluate it. The proposed system can be used to collect operating signals regarding mechanical wear-out of parts and can allow manufacturers to track state of wear and send warnings to tool owners before wear-out.