近年來智慧農業的實現，不論是感測數據的分析、無人機巡田後影像上傳的處理、乃至與物聯網的農業相關應用或農業自動化的實現，最終多數仍需仰賴雲端服務的幫助，然而雲建置的成本高昂，絕非多數企業與農民所能負擔。即便有些現有的雲服務可使用，如感測器模組供應商所提供的雲服務，仍需受限於供應商對雲系統的相關限制，缺乏彈性。 本論文提出以無伺服器運算(Serverless)為開發核心的雲架構，能有效降低企業及農民佈建智慧農業雲端系統的成本，取代傳統伺服器架構的雲服務，除了省去傳統伺服器與網路設備及相關軟硬體購置與佈建的費用，也降低了雲佈署的難度與後續維護的維運成本，且大幅提升了系統的收縮彈性與高可用性，在實際與LoRa node農業感測器搭配佈署並針對其使用成本進行演算後，實驗結果驗證了其適用性。 此種模式的雲架構並不受限於農業運用，也適用於任何以事件而觸發運算的情境，例如: 使用者點選軟體應用程式上的功能、互動式機器人等；此種抽象化式的雲架構佈署，提供企業與農民一個佈署成本或使用雲服務時，成本更低廉的智慧農業雲端系統架構。 ;Recently, the realization of smart agriculture, whether it is data analyzing, or image uploading to the cloud after drone tracking, or even the implementation of agricultural related applications or automation, ultimately are all utilizing cloud services. But the high cost of cloud implementation is not affordable by most enterprises and farmers. Even if some existing cloud services are available, such as the sensor module suppliers, the service is still constrained by the cloud system feature limitations and lack of flexibility. This paper proposes a cloud architecture with serverless as the core development, which can effectively reduce the cost of enterprises and farmers deploying smart agriculture cloud systems to replace the existing traditional server architecture services. The lower cost of hardware and software implementation also reduces the difficulty of cloud implementation and the subsequent maintenance cost. In the meantime it also improves the flexibility and usability of the system significantly. After the deployment and the cost reduction of its use, the experimental result of the LoRa node agricultural sensor does help improving farmer’s crop productivity. This model is not limited to agricultural applications, but it can potentially be applied to other areas, such as: software applications, interactive robots, etc., to improve business performance and productivity.
