智慧農業4.0領航產業共通技術之研發

字體大小:A- A A+

單位商業會員附件下載:
計畫名稱: 智慧農業4.0領航產業共通技術之研發
計畫主持人: 呂椿棠
共同計畫主持人: 林韋至;林達德;蔡燿全;溫博浚;魏駿勝;邱奕志;郭寶錚;歐陽盟;郭鴻裕;陳世芳;柯建全;楊朝旺;張仲良;楊滿霞;陳永裕;申雍;盛中德;歐陽彥杰;郭彥甫;李龍盛;沈煜棠;艾群;吳昭正;許政穆;陳享民;林慧玲;黃振康;黃文祿;連振昌;陳玟瑾;黃膺任;陳建德;洪滉祐;謝慶昌;莊益源
計畫編號: 106農科-18.2.1-子-C1
計畫主管機構: 行政院農業委員會
計畫執行機構: 行政院農業委員會農業試驗所作物組
全程計畫年: 2017
關鍵字: 機械視覺;臺灣優良農產品規範;資通訊技術;智慧農業;智慧輔具;特徵波段選擇;網實系統;雲端運算;菇類視覺分析;可調控整資料介面;影像處理;害蟲自動監測;感測器融合;微型攝影機;專家系統;物聯網;手機APP;高光譜影像;水分感測器;影音串流;資訊平台;腦波視覺控制;行動應用;無人機;無人機;數據整合;訊號傳輸;農藥殘留檢測;控制系統;非農藥防治;家禽;具整合訊息與讀寫功能之電子標籤系統;精準農業;機器學習;資料庫及雲端;無線射頻驗證;智能人機輔具;QR編碼;水產加工;茶;高光譜即時演算法;嵌入式系統;生產履歷;Machine vision;Taiwan good agriculture practice;Information and Communication Technology;Intelligent Agriculture;Assistive robotic exosuit;feature band selection;Cyber Physical System;Cloud Computing;Vision analysis for mushrooms;Adjustable raw data inter
摘要: 智慧生物感測共通平台技術研發與農業應用 整合農業環境之影像資訊及環境資訊,運用嵌入式系統之高速運算與通訊能力,同時結合感測器融合方法、isPLC場域控制技術、雲端與行動運算、以及機器學習演算法等先進技術,以建置適用於多種農業生產環境之生物監測的智慧生物感測共通技術平台。 所開發完成的平台可應用於設施蟲害偵測與分析、設施作物栽培作業管理、禽舍環境與動物行為監測、作物生長監測、果樹與田間蔬菜作物監測等廣泛之農漁牧產業應用場域。本單一計畫下共規劃有8個工作項目,透過跨領域與跨研究團隊合作分別進行智慧生物感測共通技術平台的各項核心技術開發、測試與驗證。有關共通技術平台核心技術開發的項目包括:1.整合型生物感測模組技術開發;2.嵌入式系統感測模組與isPLC場域控制之研發;3.感測資訊之通訊協定與感測器融合技術;4.雲端資料整合及行動裝置應用技術開發;5.影像資料應用端資訊分析及機器學習演算法技術開發等項目。所開發的共通技術平台將實際於溫室管理、家禽生產管理、蟲害自動監測、果園與田間蔬菜作物監測等多項農業應用場域進行測試與驗證。精準/無農藥殘留茶園栽培管理技術開發    為促成茶商與契作茶農建立「智農聯盟」的合作夥伴關係,本計畫擬於2年研究期程中就(1)開發可以偵測茶菁品質與產量空間變異分布的遙測技術,以及繪製契作茶園精準肥培管理組圖,協助茶農生產品質均一的茶菁;(2)開發可以依據茶樹品種和氣象觀測資料,推估最適採摘日期的統計預測模式,以有效協助茶商/茶農調配不同茶園茶菁採摘時程和採茶工人力;(3)研發茶園病蟲害發生密度預測模式和非農藥病蟲害防治技術,以保障消費者健康,並提升品牌的知名度和售價等3項關鍵技術進行研發。預期可以達成維持品牌茶葉風味的穩定性與供貨量,提升消費者信心和購買意願,擴大臺灣茶葉國際市場通路,以及增加並穩定茶農收益等效益。高光譜農業4.0前瞻研究:高光譜即時農產品品質監控平臺技術-以蝴蝶蘭和杏鮑菇為例    本計畫希望藉由高光譜成像技術,改善現有檢測方法,建立標準化、非破壞性且即時的農產品品質檢測平臺。 該檢測平臺如結合農藥之高光譜資料庫,亦可應用於農業蔬果的農藥檢測,將可大幅提升食品安全性的保障。本計畫預計將執行兩種實驗標的物,分別為杏鮑菇及蝴蝶蘭,所預進行包含杏鮑菇含水量及病兆檢測以及蝴蝶蘭切花後壽命預測和黃葉病兆檢測,以達成驗證高光譜檢測平臺使用之可行性。透過所架設之高光譜檢測平臺測的實驗標的物(蝴蝶蘭及杏鮑菇)之光譜數據,首先建立實驗標的物在不同狀態下之資料庫,透過建立高光譜單點及影像數據資料庫,對待測物反應較敏感的波段進行提取及分析,並且利用光譜解混合(spectral unmixing)演算法,將待測物之品質控管程度量化。建立以高光譜檢測農產品質監控之平臺,可藉由檢測達成全面性的產品品質管控及產量提升,達到優質農產品之優質外銷,建立臺灣國際形象,也可提高農民收益及建立消費者對農產品的品質信心。藉由自動化量化檢測結果,可對農產品品質進行全面的管控。生物特徵辨識系統之開發-以種鵝為例    主要以影像技術為基礎,發展國內畜禽產業適用的以外觀特徵為主的身分辨識技術及疾病階段之行為特徵辨識技術,以種鵝為主進行開發,完成開發所需的種鵝特徵偵測及特徵參數擷取試驗平臺之建置,及試驗軟體開發。利用影像技術開發人工ID自動識讀軟體,可取代價格較昂貴的電子標籤及識讀系統,並探討傳統人工標誌於種鵝身上作為身分辨識,如刺墨、色塊、烙印編號、懸掛身分證、條碼標籤等方式應用於種鵝辨識的適用性,評估分析及選訂可適用於種鵝的人工ID,可輔助進行個別鵝隻之身分快速辨識並辨識禽舍、所在位置及運動狀態等資訊,以發展更精確及更快速的生物特徵辨識系統,有效提升種鵝身分自動辨識率,所蒐集的大數據才能真正說明動物的狀況,進而用來改善動物的飼養管理效率,降低種鵝生產管理作業及追蹤追溯所需的勞力與建置成本。智能化露地田間感測訊號傳輸與系統之開發    目前露地契作地瓜田之土壤水分,皆以巡田人員攜帶土壤水分感測器,進行現場測試,且無適用遠端廠區資訊平臺之監控管理系統,對地瓜生長狀態、雜草及病蟲害皆應用人力巡田,耗用人力大。有鑑於此,本計畫主要目標在研發智能化露地田間狀況之管理,所需的即時供水控制模組、建立遠端監控資訊平臺之管理及應用無人機進行契作地瓜田之生長狀態、雜草及病蟲害影像監控管理平臺。透過本計畫之研發成果,未來可比照應用於其他露地農作物智能化田間狀況之管理,以降低田間管理人力,確實且精準掌控田間狀況,提升智慧農業技術及農民利潤,提高永續經營農業的競爭力。農牧業生產用水智能感應器之開發    主要在引入日本水稻田傳感器PaddyWatch及其管理策略,可根據水田高低水位的設定進行灌溉管理,並根據家禽產業的水質管理需求,開發國內牧業生產適用的用水之智慧化管理感應器之試驗機型,並開發應用程式及智能型農用感測器共用的共通平臺,於家禽飲用水品質進行監測試驗及自動化收集用水數據。未來所開發的感測器與系統將實地於水稻田及養雞場進行水資源利用及飲用水品質之監測試驗,自動化收集生產用水數據。本研究成果可有效的提昇農牧業生產用水管理精準度,提昇勞動效率,並提高生產品質及降低生產成本等效益,亦可配合未來建立的智慧農業4.0共通平臺,建構國內農漁牧業生產環境、水資源利用及水質管理之大數據自動化收集、整理及上傳平臺應用系統,進行分析及加值應用,達到水資源的最佳化利用及確保用水衛生安全,且所收集數據亦可供為未來追蹤追溯用,提升食品安全及消費者信心。動態生產履歷系統-以遠洋漁業為例    完成漁獲可追溯系統之資料寫入電子標籤與生物相容性設計,並建構使用者端可視化介面平臺與生產端溯源資料庫,先期研究將以高單價大型漁類(如鮪魚)為主要對象。第1年之詳細技術內容包含開發具可即時寫入之具生物相容性電子標籤系統(包含產地位置、產品數量、大小及長度等),以及建構生產端漁獲溯源管理系統介面、資料庫與可搜尋化網頁平臺。第2年則整合與測試漁獲可追溯系統之資料即時寫入與使用者端溯源管理系統,並進行使用者端漁獲資料遠端上鏈及傳輸功能設計。透過溯源履歷,不僅能確保漁獲來源的正當性,降低食安問題的發生率,且消費者在選購漁獲時,能夠由「漁獲溯源履歷平臺」得知該漁獲之清晰透明資訊,包括漁獲生理訊息、物流環境訊息及參與人員的誠信紀錄等。藉此期能讓消費者不再盲目依從價格購買漁獲,而是選擇更明智與更環保的購買方式,並使消費者對國內漁獲產生信任感,提高國內漁獲帶來的經濟效益。而所有的漁獲訊息將永久保存於漁獲資料庫,一方面能使管理者藉此了解消費者的購買習慣趨勢,另一方面可進行捕撈至販賣的路徑分析,以減少物流成本,提高生產者收入。希望透過建立漁獲溯源履歷系統,營造一個消費者、生產者以及管理者三方共利的局面。另開發完成之電子標籤系統,因其具備輸入訊號之可調整性,故僅需簡單之軟體設定即可搭配不同物品感測裝置(如影像感測器、重量感測器等),便能進行各式產品或物品的即時動態標籤輸出與資料上傳雲端網路資料庫儲存,此可攜式電子標籤系統將讓現存已建構相當完整之農產履歷回溯系統,於履歷電子標籤化的過程減少標籤設備或產品的搬運與人力的使用。物聯化水產品初級加工處理省工機具之研發    在水產加工處理流程中,魚類的初級加工處理(三去三清)大多需仰賴大量人力及資源投入加工處理流程,然而初級加工處理操作重覆且固定然而操作人員經驗往往決定生產效率及取肉率,因此搭配智慧化物聯網系統、環控感測技術及自動化加工處理省工機械輔助即為提升加工處理效率的重要方向,藉由自動化低成本加工處理省工機具除可提升生產效益,並減少由人為造成之加工處理生產耗損,更可提升水產品加工流程標準化及食品安全,促使加工水產品得到更優質的保障。穿戴式採收和搬運智慧人機輔具之開發與應用    本計畫將開發一智慧型採收與搬運之個人穿戴機構,藉由輕量化高強度之輔具來增加搬運過程中的助力以及減少重複動作所造成之傷害,並透過與手機app以及物聯網的結合,發展出1套作業系統,不僅能提供新一代農人更簡便的方式管理農場及農作物的採收情形,也能透過此系統實現智慧農業4.0之技術研發。本計畫之主軸為開發一採收搬運之智能人機輔具,計畫內容將細分為農業輔具、物聯網系統以及體能感測系統3大架構。在農業輔具的部分,將著重於省力輔具機械設計開發與客製化智慧穿戴式機構,首先挑選以輕盈堅固為特點之材料應用於機構設計中,例如碳纖維材料等,在分別針對此穿戴式機構的舒適性、萬向支撐結構、驅動與平衡系統、智慧化自動控制系統以及感測器回饋系統進行研究開發。在物聯網系統中,將自主開發出一農產品智能採收與管理系統,以及利用樹莓派作為通信設備進行資料的傳輸,透過與網際網路的整合,將傳統產業結合新一代技術並吸引年輕世代回流。而在體能感測系統的開發平台上,將偏重於視覺分析以及視覺控制,透過輔助裝置來進行農作物的挑選,針對各式菇類之方位影像偵測與成熟度分析,以即時回饋至智能輔具採收系統上,來快速達到高採收效率。本計畫所開發之智慧農業輔具,將利用其獨創性以及實用性,在技術創新成就上有所突破,藉由輔具結合物聯網系統之智慧穿戴式機構,提升傳統生產機械的效率。而在經濟效益部分,透過智慧輔具幫助農民進行生產作業,將大幅改善人力成本的支出,進而壓低農作物的市場價格,提高與進口商品的競爭力。動態生產履歷系統建置-有機農產品生產為例    近年來農產品安全問題頻傳,已引起政府與消費者高度重視。為了提升產品品質、落實永續農業發展,政府及相關業者希望發展能結合臺灣農業規範及產銷履歷之系統,讓農產品由田間生產管理至消費者之餐桌間(from Farm to Table) 的所有流程均有可追溯之紀錄,也可使產品生產過程能更公開、透明。消費者也能透過智慧型手機或網路即時查詢產品各生產階段的狀況,但現有系統之整合性不足,且都採較靜態的方式呈現。有鑑於此,為能有效展現產品在各階段的生產狀態,讓消費端能了解產品在各階段的生產狀況,本研究提出具有產銷履歷回溯與查詢功能,且結合產品生產時之動態影片,發展出「動態產銷履歷系統」,此系統將建置產銷履歷、微型影像記錄之攝影機、影片資料庫、自動影像辨識、雲端伺服器、管理系統、QR Code產生與查詢以及手機APP程式等系統,讓整個供應鏈之生產者、銷售供應端、消費端等都能有效存取相關資訊。另為確認系統使用效能,本研究將與有機產品業者合作,進行各種系統功能試驗與測試,不僅生產資訊紀錄能更透明,也能建立消費者對產品食用信心,藉此增加產銷履歷產品之附加價值。智慧農業4.0領航產業共通技術之研發一、高光譜即時農產品品質監控平台技術開發:針對影響瓶插壽命的因子,如吸水性、乙烯敏感度及內部養份含量等建立文心蘭切花非破壞性品質檢測模式,同時建立杏鮑菇含水率非破壞性品質檢測模式。本計畫目標為建立大量農產品高光譜資料庫以用於分析農產品品質,以及尋找反應較劇烈的特徵波段,以提高高光譜分析的準確率。 二、農業無人機的監測與噴灑技術應用於作物病蟲害防治與養分、水分管理:完成引進與改進1人多機或單機操控農用多旋翼噴藥、葉面施肥無人機(UAS)應用及農用無人機的標準作業程序;建立針對各種作物生長期、病蟲害的防治(不同用藥之霧粒粒徑選擇)與葉面施肥(濕潤時間長)需求噴藥方法與防治效果評估;農用無人機專用肥料開發評估與現行農藥效果評估。 無人機多源感測器農作監測系統應用研究。 三、 建構農業害蟲智能監測暨管理決策系統 :研發自動辨識及計數的智慧監測裝置,透過系統自動辨識害蟲及計算數量,進行雲端巨量分析與數位決策;重要農作物有害生物鑑定及智能查詢系統架構; 針對蔬果糧作及倉儲研發之病蟲害管理技術策略為主要內容,於共通資訊平台上建置以作物或害蟲類別為主體個別作害,第一年預計於測試平台上建置6項重要作物/害蟲之整合管理策略共通性資訊平台運用農業生產力大數據水庫(Data Lakes),透過知識參數與概念關連擷取整理,將專家知識參數化並建模,形成專家群決策模型,並依事件定義驅動農業生產流程以及參數化彈性調整機制,透過回饋機制動態的調整流程,修正農業生產的決策模型。主要發展內容為「農作物栽培知識庫」及「農業技術參數管理技術」發展建立「農業大數據分析」與「大數據資料管理」,收集氣候、土壤肥力、聯網設施、地理環境、作物生產流程等資訊,提供設施農業生產所需之管理決策建議,及環控設施之管理模式建議。Research and Development of Intelligent Bio-sensing Platform and Its Agricultural Applications    The goal of this project is to develop an intelligent bio-sensing common platform for agricultural applications and production sectors. The common platform integrated the advanced technology including embedded system with high speed computation capability and communication functions, sensor fusion, isPLC field control, cloud and mobile computing, and machine learning. The integrated common platform can be applied in wide applications in agricultural sectors such as pest monitoring and control in greenhouse, crop cultivation management, poultry facility management and animal behavior monitoring, plant disease detection and early warning, orchard and field vegetable monitoring, etc. This integrated project is composed of 8 sub-projects with multidisciplinary researchers from various research institutions. The aim is to develop core technologies of the bio-sensing common platform and deploy them to real agricultural applications such that the developed system can be tested and validated in field. The core technologies of the common platform include: 1. Integrated bio-sensing module with image and environmental sensors; 2. Integrated sensor module and isPLC field control using embedded system; 3. Protocols of sensing data and sensor fusion technology; 4. Cloud integration platform for a universal data management and mobile device application development; 5. Image analysis and machine learning tools. In the process of technology development, we will test and validated the integrated bio-sensing common platform in collaborative sub-projects for various agricultural applications including greenhouse management in crop cultivation, poultry facility management and animal behavior monitoring, greenhouse insect pest monitoring, orchard and field vegetables monitoring.Development of Precision and Pesticide-Free Management Techniques for Tea Plantations    Tea (Camellia sinensis (L.) O. Kuntze) is an important economic crop in Taiwan. Green tea, black tea and Oolong tea are major types of teas produced in Taiwan. Among them, Oolong tea is best known to the world. In order to facilitate the formation of smart alliances between tea merchants and contract tea farmers, this 2-year project intents to develop three key technologies: (1) mapping spatiotemporal variations of growth within tea plantations by remote sensing techniques, and producing site-specific nutrient management maps for contract plantations to stabilize the quality of plucked new shoots; (2) statistic models for predicting best harvesting date based on varieties and weather records to assist adjustments of labors and equipments during the period of new shoots plucking; (3) pesticide-free cultivation practices to protect consumers’ health, raise brands’ reputation and market prices. It is believed that this project has the potential to maintain a stable supply of good quality raw materials for making brand name teas, to raise consumers’ confidence and buying intentions, expand international trade of Taiwan teas, and increase tea farmers' incomes.Agriculture 4.0 on Real-time Vegetables And Fruits Inspection by Hyperspectral Image Analysis and Test Platform-Take Phalaenopsis and Pleurotus As an Example    With the improvement of people's quality of life, the quality of agricultural products gains more and more attention from consumers. The quality of agricultural products is the key to protect the product value, increase economic efficiency, however, the current quality control of agricultural products is to detect more by virtue of manual way to determine the quality of the product. Manual way of detection is not only time-consuming but also lack of objective basis, and hence cannot observe problems from the appearance and more likely miss leading of quality control. To improve the existing detection methods by using hyperspectral imaging technique and to establish a standardized, non-destructive and immediate agricultural product quality examining platform. The examining platform, such as the combination of pesticide hyperspectral database, can also be applied to agricultural and vegetable pesticide testing, will be able to significantly enhance the safety of food safety. The proposal is expected to carry out two kinds of experimental objects, namely Pleurotus eryngii and Phalaenopsis. The experiments will be conducted to test Pleurotus eryngii for water contain and disease detection. We will also test on Phalaenopsis for cut flower life prediction and yellow leaf disease. Those examine results could lead us to achieve the verification of the feasibility of using the hyperspectral detection platform.  Based on the spectral data of the experimental objects (Phalaenopsis and Pleurotus eryngii) measured by the hyperspectral detection platform, the data of the experimental subject in different states were established. Through the establishment of single point and image hyperspectral database, the sensitive part of the response to the measured part of the extraction and analysis, and the use of spectral unmixing algorithm, the quality of the test object can be control by the degree of quality. The establishment of hyperspectral detection of agricultural quality monitoring platform, through the detection of a comprehensive product quality control and production to achieve high-quality agricultural products hence can have high-quality export and the establishment of Taiwan's international image. At the same time we can also increase the farmer’s income and to establish products quality confidence for consumer. Through the platform, test data can be automated examine and classified. The quality of agricultural products can have a comprehensive control hence the goal of food safety can be reached.Development of a Feature Recognition System for Life-Breeding Geese as a ExampleThe project in this year will focus on the application of machine vision to identify the apparent characteristics of breeding geese. The examined platform with machine vision function and the operating software will be developed and established. The developed animal ID recognizing system with machine vision can be used to replace the expensive the electronic tag and identifying system. But the big animal for example cow can still use the electronic system to identify the animal itself. In addition to examining the biological characteristics of animal, the application of artificial mark to identify breeding geese will also be explored. For example, tattoo, color block, branding, hanging id, barcode and matrix id, etc. The id technique will be fully studied and evaluated to be used in poultry industry. The developed will be further emphasized to identify the animal¸ barn, location, and its activity. The developed system can be more accurate and faster to identify the animal with biological characteristics. It can identify the animal easily and successfully, then the collected data can really illustrate the animal situation and meaningful. Then the big data can be used to improve the management efficiency. The developed system can really reduce the labor load and cost to establish the managing, tracing and tracking system.The Development of Intelligent Detective Signal and the Transmission System in the Outdoor Field    At present, the soil moisture of the open land in the sweet potato fields just depends on the patrolling field personnel to carry the soil moisture detecting device, to carry on the spot test and not have applicable platform to carry remote factory information control administrative system. The growing state of sweet potatoes, weed or the plant diseases such as insect pests should be employed and consumed by a lot of the manpower to patrol the field. In view of the situations, this plan main goal is intelligently to Research &Develop the open land of the management of the ground field state such as the necessary one supplies water of controlling the model group immediately, setting up and carrying the management which controls the information terrace and employ the unmanned aerial vehicle to carry on the contract and detect the growth state of sweet potatoes, weed , plant diseases or the insect pests etc. which can be imaged in the remote factory information control administrative system. Based on the R&D achievements, it can be applied for the similar situations such as the intelligent management of open land or field detecting in the future. It means that the achievement can reduce the manpower of field management, control the field state certainly and perfectly, improve agricultural technology of the intelligence and peasant's profits and improve the competitiveness of managing the agriculture continuously forever.Development of Intelligent Water Sensor for Agriculture and Animal Husbandry    In recent years, the rapid changes in climate lead to water shortage or too much, however, domestic agriculture uses nearly 70% of the water resources. The water resources more and more precious in today, how to use the existing technology, to optimize the use of water resources should be a serious challenge in the future. Coupled with rural human aging and lack of serious problems, and the water resources and water hygiene management need to spend a lot of labor, making the urgent need to introduce and develop domestic agriculture and animal husbandry intelligent water management common technology and equipment. In this year, the PaddyWatch and its management strategy were introduced. Irrigation management can be carried out according to the setting of high and low water level of paddy field. And according to the water quality management needs of the poultry industry, development of suitable for the use of water intelligent management of the sensor test machine for domestic animal husbandry production,And develop apps, to monitoring the quality of poultry drinking water and automatic collection of water data. This study can effectively promote the precision level of water management in agriculture. The developed system can be fully linked to the common platform of intelligent agriculture. The system, which and automatically collect and transfer related data to the common platform, will be used to establish the intelligent production environment, the big data of water quality and resource management. The collected data will be further analyzed and evaluated to create the extra value application. The developed sensor and system can be used to optimized the water application and management, and also ensure the water safety. The collected data can be used for tracing and tracking, which can further promote the food safety and consumer confidence.Design of Dynamic Traceability System for ocean fishingThe existed traceability system delivered really unreliable products to buyers due to no effective traceability systems are issued in the past decades hence it is almost impossible to track the hazardous fish products immediately when some of them are confirmed as being harmful to consumers. The successful development of this proposed project will provide an entirely new traceability system for the fish catch process which still has no an effective treatment in dealing with the trace of the fish catch. The developed and advanced traceability system offers the possibility for consumers to precisely trace the origin and trajectory of catches sold in the market , and this is a totally different character of this proposed design with respect to the existing traceability system.The overall design target of this project in 2018 is to develop a traceability apparatus, an signal communication interface and three data bases and websites for the purpose of tracking the traces of catches of ocean fishing.  No doubt, the successful development of this project will guarantee: 1. The qualities of all legal catches for all consumers, 2. Cost reduction of the logistics and 3. The management unit or supervisor can monitor the habits of consumers buying catch products by using the storage data of  this proposed traceability system. Furthermore, this proposed traceability system offers the ocean fishing companies a really definite standard to comply with. The achievement of this electrical traceability apparatus can be applied to many kinds of products in our normal life for recording purposes due to the flexible design which is with the help of adjustable interface of multi-signals and the tunable sensor fusion ability.The research for the Internet of things seafood primary processing with labor-saving machinery    In the aquaculture process, the primary processing of fish relies on a large amount of manpower and resources to do process, whereas primary processing operations are repetitive and fixed. However, operator experience often dictates productivity and meat harvesting .Therefore, with the wisdom internet of things system, environmental sensing technology and automated processing of labor-saving machinery that is to enhance the efficiency of processing an important direction, with low-cost automated processing equipment in addition to provincial workers can improve production efficiency and reduce production wear caused by the human, but also enhance the standardization of aquatic products processing and food safety, to promote processing of aquatic products to better quality.Development of assistive robotic exosuit for agricultural applications    Crops collection and carrying rely on dense labor recently. How to effectively reduce the cost, raise the market competitiveness, increase the traditional industry to a new level which is close to human life have become big issues. It not only provides farmers a more convenient system for management but also raises the output value of domestic crops. Therefore, a smart collection and handling of stand-alone assist device will be developed in this project. Wearing assist suit which is lightweight and high-intensity will not only increase the effect of handling process, but reduce the damage from the repetitive actions. To provide a more convenient way for managing the farm and the harvest of crops, assist device will arrange a pairs with mobile APP and Internet of Things (IoT) which can also develop a technology of Agriculture 4.0. The main purpose of this project is to develop a smart agricultural assist device for collection and carrying. The project is divided into three parts, which are agricultural assist device, internet of things, and physical sensing system. In the first part, we focus on the development of strength-saving mechanism design and customized wearing device. The light and strong material such as carbon fiber is applied to the mechanical design. The characteristics of the wearing device are optimized and developed based on the comfortability, supporting structure, driving and balance system, smart automatic control system. In the internet of things (IoT) part, we develop a managing system for smart collection and take advantages of Raspberry Pi as a communication device for data transmission. Through the internet integration, the traditional industry is combined with new generation technique and attract more young people to pour themselves into this field. On the physical sensing system platform, it will focus on visual analysis and visual control and use this system to collect crops by the assist device. By the video detection and maturity analysis of mushrooms, it will carry out instant feedback to raise the efficiency of collection. The developed smart agricultural assist device with Internet of Things (IoT) system will be applied on crops moving and information of market transmission. On the one hand, the farmers can not only reduce the burden during busy farming season but also realize the requirement of custom directly without businessman. Therefore, the farmers can maintain the quality of crops and receive appropriate remuneration. On the other hand, by setting up a research group and design the assist device, the researchers will learn the ability of independent thinking, how to make good use of the current source and bring into the teamwork spirits. The research works are written into international paper and submit to mainstream conference and journal, and we expect that this smart agricultural assist device can be turned into a spotlight by discussing with specialists in different field.Building a Dynamic Production Traceability System-Production of organic agriculture Product as Example    Recently, safety of agricultural products has attracted government and consumers. In order to increase product quality and result in permanent development of agriculture, government and business hope to develop a system which can integrate Taiwan Good Agriculture Practice (TGAP) and Production Traceability System. The agricultural products from the production management of farm to consumer's table can be traced completely with various records. That is, the product manufacturing process can be more opened so that the factors affecting qualities can be handled. Moreover, consumers can the production status of product via smart phone or internet. However, the current developed systems still do not integrate these two systems, and developed in a static way. To this end, the research propose a system which can integrate these two systems with query and traceability ,and effectively display the production statuses in various production stages so that consumers can easily realize the production statuses for their bought products. Above all, the proposed system combine the production videos in order to develop a "Dynamic Traceability System" including such as production and sale traceability based on radio frequency identification (RFID), camera with micro image recording,  video database, automatic image recognition, clouding server, management system, QR code generator and query, related mobile APP. The entire supply chain can be linked including manufacturers, suppliers, consumers with various access functions.In order to validate the performance of the proposed system, we will cooperate with organic agriculture business to test the developed functions. We believe that if the proposed system can be developed successfully, the production information will be more opened and transparent, the consumer's confidence can be enhanced, and the extra value of the agricultural products can be increased finally.Development of common technology of smart agriculture 4.0 for leading industries      Development of the real-time monitoring platform of agricultural product quality with hyper-spectral technology: for the factors affecting the life of the vial, such as water absorption, ethylene sensitivity and internal nutrient content, the establishment of Oncidium cut-flowers testing mode, at the same time establish the Pleurotus erygnii non-destructive quality testing mode. The purpose of this study is to establish a database of agricultural products for the analysis of the quality, and to find more responsive characteristics of the band to improve the accuracy of hyper-spectral information analysis.       Application of UAS surveillance and spraying techniques for crop pest control and nutrient and water management : to introduce or improve an automation architecture for single operator, multiple UAV command and control mode of agricultural multi-rotor spraying, foliar fertilization UAV(UAS) applications ;to compile an agricultural unmanned aerial vehicles standard operation procedures in the field; to establish an assessment methodology interms of the control of pests and diseases (different treatment of spraying fog particle size selection) and foliar fertilization (long wetting time)requirements in the different crop growth phases; to develop a specified fertilizer for UAV and assess market generic pesticides effect. And to study on the application of UAV multi - source sensors in agriculture monitoring system.        Construct intelligent pest monitoring and management platform for pests of agricultural importance :Major working items include develop automatic pest monitoring devices, risk assessment technique and decision-making system for thrip and whitefly; establish rapid on-line pest diagnosis platform and provide related control tactics and strategies based on successful field implementation. The biology of major pests on fruit, vegetable and storage grain, their damage patterns, and individual management strategy will be constructed in the platform. Details of integrated pest management on six important crops or pests will be accumulated and proposed in the first year.Common Information PlatformBy using Data Lakes, the knowledge parameters were correlated with the concepts, and the experts' knowledge was parameterized and modeled to form the expert group decision-making model, and the agricultural production process and parameterized elastic adjustment mechanism were driven by the event definition. , Through the feedback mechanism of dynamic adjustment process, modify the agricultural production decision-making model. The main development contents are "Crop Cultivation Knowledge Base" and "Agricultural Technology Parameter Management Technology". The establishment of "Agricultural Data Analysis" and "Big Data Management" collects climate, soil fertility, networking facilities, geographical environment and crop production processes. Information to provide facilities for agricultural production management decision-making recommendations, and environmental control facilities management model recommendations.
資源連結: 前往查看