精準農耕技術導入水稻灌溉水遠端監測之使用者經驗與雲端數據分析

User Research and Cloud Data Analysis on Remote monitoring of Rice Irrigation Water using Precision Agricultural Technology

陳舒玲(Chen Shu-Ling)

指導教授 ： 陳聖智 ； 蔡銘峰

國立政治大學/傳播學院/數位內容碩士學位學程/碩士(2019年)

https://doi.org/10.6814/THE.NCCU.DCT.005.2019.B02

摘要

隨著資通訊應用普及、農政力行推廣，臺灣農耕環境已逐漸從機械化、自動化導向智慧化生產管理。本研究以Norman (2013)提出的人本設計流程的觀察、衍生想法、製作原型及測試流程，探討資通訊技術應用於臺灣農作之大宗－水稻的灌溉水管理，以人本設計流程洞察目前水稻栽種的潛在需求、開發系統並導入農家驗證。本研究分為以下三階段：第一階段為需求探求階段—以個案參與觀察初探水稻耕作環境，再以半結構式深度訪談農友，發現兩項可做為研究導入技術應用的需求：(1)農忙期的工作調度雙向溝通；(2)灌溉水情異常立即得知與管理決策，後者為水稻生長與農家巡田之重點觀察，因此以後者為研究方向。第二階段進行系統開發與建置：設計物聯網遠端監測灌溉系統AgritWater，以Arduino開發農田感測器、架設數據儲存及提供自動化推播的雲端伺服器及提供資訊呈現的App平台，設計即時、遠端及精準的灌溉監測系統。第三階段將系統導入農家實驗：系統導入農家體驗七日，以深度訪談、評估量表及雲端使用者操作紀錄綜合分析，驗證灌溉水遠端監測的需求及可行性。研究成果顯示：系統有助於農家灌溉水遠端監測、降低巡田次數，系統數據有助於輔助農家決策灌溉水排灌管理的參考，整體系統使用性(SUS)平均加權分數為82分，為農友可接受的設計。在彙整使用者對數據的解讀，系統設計建議可從農民耕作管理決策判斷，規劃智慧服務應用於生產決策，結合氣象與農地環境數據建立常態模型，將系統服務導向智慧灌溉管理的決策應用，提供更精準的農情資訊推播，使推播資訊更契合農友實務耕作經驗及決策情境。

關鍵字

智慧農業 ； 稻米生產 ； 數位內容與科技應用 ； 物聯網 ； 資通訊科技 ； 灌溉水 ； 精準農業

並列摘要

With the widespread of information communication applications and promotion of agricultural policies, Taiwan's farming environment has gradually shifted from mechanization and automation to intelligent production management. This research, based on observation, idea generation, prototyping and testing procedures of human centered design process proposed by Norman (2013), explores the information communication application of irrigation rice water management in Taiwan. During the design process, potential needs of rice planting are observed, the monitoring system of irrigation water is developed and verified by farmers (during period). This research has 3 stages: demand exploration, system development and experiment. The demand exploration stage applies participant observation to study rice farming environment and context on a case-by-case basis, then interviewed farmers using semi-structured in-depth method. Two approach that is suitable for application research is discovered: (1) Two-way communication work scheduling during busy period; (2) Real time irrigation water abnormality notification and management. The latter is chosen for research because irrigation water is the key observation for rice growth while farmers patrol. The system development stage designs the real-time remote precise monitoring irrigation system AgritWater. The AgritWater uses Arduino to develop the sensing device, data collected are upload to cloud servers simultaneously and sync to applications platform for information representation. In the experiment stage, AgritWater is introduced to the farmers, the farmers then spend 7 days using it to monitored their irrigation water. The requirements and feasibility of AgritWater were verified by comprehensive analysis of in-depth interviews, assessment scales and cloud user operation records. Experiment results show that farmers benefit from the system and reduced the times of inspections. The sensing data also support farmers in decision making of irrigation water drainage management. The overall system usability score (SUS) is 82 points, proven the AgritWater is an application acceptable by farmers. In consolidation of user data interpretation, it suggests the system should provide decision making management service. By integrating weather and field environment data to construct a constant model for the environment, the system then can provide a much more precise information on field condition, which is more relevant to practical experience of farmers and their decision making context.

並列關鍵字

Intelligent agriculture ； Rice production ； Digital content & technology application ； Internet of things ； Information and communications technology ； Irrigation ； Precision agriculture irrigation

參考文獻

中文文獻

八木宏典、周若珍(譯)(2015)。超圖解米的基礎知識。臺中市：晨星出版有限公司。

天野英二郎(2014)。推動智慧型農業-ICT與自動化裝置之應用。立法與調查，第359期，頁44-56。

申雍、吳正宗、陳仁炫、江志峰、劉禎祺、王鍾和(2003)。應用遙測影像輔助判釋土壤限制因子分佈地區之技術。水稻精準農業體系，第105號，頁81-88。

行政院農業委員會(2016)。農業統計年報104年。臺灣：行政院農業委員會。