摘要

本文將審視當今製造業麵臨的核心挑戰，探索正在席卷行業的變革浪潮。這場變革源於對資源敏感型製造的全新關注
，而人工智能、分散式控製、混合組網及軟件定義自動化等新技術與能力協同發力

，共同為未來數字化工廠的崛起築牢根基。

製造業麵臨的挑戰

zhizaoyezhengchuyuyichangzhuanxinglangchaozhizhong

，xiaofeizheduigexinghuachanpinxuqiudezengchang，jiazhiyiqinghougongyinglianweijicuishengdechanyehuiliuqushideng，chengweituidongzheyibiangedezhuyaoqudongli。erzhexie，jinjinshizhongduotiaozhanzhongdebingshanyijiao。yucitongshi，quanqiugeguozhengfuyefenfenchutaixiangguanfagui，yijianshaozhizaoyedetanpaifang，congershixianwenshiqitijinglingpaifangmubiao。yingduizhexietiaozhanjiangweigongyezhizaoqiyekaipiquanxindefazhansaidao，qiyekejieciqijiyinruqianyanjishu，zaijiangditanpaifangdetongshi，tigaozhizaoyedeshengchanxiaolv、可擴展性和靈活性。

zairujinjiyoudezhizaogongchangnei
，zhizaoshebeiyuzidonghuashebeilijingduonianfanfubushuyukuozhan，hucaozuoxingwentiriyituxian。shebeijianbujinnanyishunchangxietongyunzuo，xianghujiandelianjieyejiweiyouxian，daozhigongchangneibupubianquefanengguantongsuoyouzidonghuashebeidetongyiwangluo。

隨著新產品庫存單位(SKU)數shu量liang持chi續xu攀pan升sheng，生sheng產chan線xian的de設she置zhi與yu驗yan證zheng時shi間jian不bu得de不bu相xiang應ying增zeng加jia。在zai醫yi療liao器qi械xie製zhi造zao領ling域yu
，驗yan證zheng流liu程cheng不bu僅jin耗hao時shi漫man長chang，成cheng本ben也ye十shi分fen高gao昂ang。此ci外wai，產chan品pinSKU的增多還會拉低設備綜合效率(OEE)，原(yuan)因(yin)在(zai)於(yu)額(e)外(wai)投(tou)入(ru)的(de)設(she)置(zhi)和(he)驗(yan)證(zheng)會(hui)造(zao)成(cheng)生(sheng)產(chan)時(shi)間(jian)的(de)浪(lang)費(fei)，進(jin)而(er)導(dao)致(zhi)生(sheng)產(chan)效(xiao)率(lv)下(xia)滑(hua)。製(zhi)造(zao)業(ye)麵(mian)臨(lin)的(de)挑(tiao)戰(zhan)不(bu)止(zhi)於(yu)此(ci)，熟(shu)練(lian)工(gong)人(ren)短(duan)缺(que)問(wen)題(ti)同(tong)樣(yang)嚴(yan)峻(jun)。據(ju)預(yu)測(ce)，截(jie)至(zhi)2030年，製造業熟練工人缺口將高達約210萬人。1 當下，多數製造活動集中於既有工廠
；zaicibeijingxia，qiyeshituzaixianyouchangfangkongjianneitishengchannengshi
，laodonglibuzudewentibianchengweichannengtishengdeguanjianzhiyueyinsu。weilaishuzihuagongchangzhengshiweigongkeshangshuzhongzhongtiaozhanersheng，zhiliyutuidongzhizaoyemairuquanxindefazhanjiyuan（見圖1）。

圖1.工業製造麵臨的挑戰。

工業製造業的轉型

從cong技ji術shu角jiao度du來lai看kan，製zhi造zao業ye已yi取qu得de重zhong大da進jin步bu。例li如ru，通tong過guo在zai製zhi造zao資zi產chan和he設she備bei上shang增zeng加jia傳chuan感gan器qi部bu署shu並bing進jin行xing融rong合he，可ke生sheng成cheng豐feng富fu的de數shu據ju集ji
，用yong於yu優you化hua機ji器qi並bing提ti高gao設she備bei綜zong合he效xiao率lv(OEE)。軟件定義自動化的部署提升了製造業的生產效率、靈活性和可擴展性
，大幅縮短了設置與驗證時間。此外，人工智能(AI)zhengzhubuxiangbianyuancefazhan，gengjiakaojinchuanganqihuozhixingqidengshengchengshujudezhongduan。bianyuanrengongzhinengjiangjiezhushujuqudongdejuecefangshi，bazhizaoshujuzhuanhuaweiqieshikexingdejianjie
，zhulizizhuzhizaoshixianzhizaoyeshengchanxiaolvyujingzhenglideyuesheng（見圖2）。

圖2.製造業的轉型。

資源感知型製造

下一代製造業需要更全麵地審視資源消耗的各個方麵。製造業所需的四大關鍵資源分別是資金、電力、cailiaoherenli。zaiziyuanganzhixingzhizaodebeijingxia，weilaishuzihuagongchangjidaitishengduizhexieziyuandeliyongxiaolv。zaizijinxiaolvfangmian，suoyouzhizaolingyudezibenzhichudouyingzhuzhongshixiantouzihuibaolv(ROI)，周期可能為一年、sannianhuowunianbudeng。weilaishuzihuagongchangdeguanjianmubiaozhiyi，bianshiyizuishaodezibenzhichushixianlirunzuidahua，jinerhuodezuigaodetouzihuibaolv。qicishidianlixiaolv，xiayidaizhizaoyebixuyigengdidenenghaoshixiangenggaodechanchu
，dachengjianshaoquanqiutanpaifangdemubiao。jiangdidianlixiaohaodeguanjianjucuobaokuo：部署高效電機驅動器，將氣動驅動替換為機電驅動，運用自適應閉環控製技術提升製造效率，等等。

ziyuanganzhixingzhizaodedisangefangmianshicailiaoxiaolv。zaitishengzhizaoyekechixuxingfangmian
，jianshaocailiaolangfeiyujiangdinengyuanxiaohaotongdengzhongyao
，fahuizhebukehuoquedezuoyong。tongguozuidaxiandudijianshaoyuancailiaodeshiyong，zaijiehejiaqiangshengchanzhiliangkongzhi

，nenggouxianzhujianshaozhenggezhizaoliuchengzhongdecailiaolangfei
，zuizhongchaozhelingfeiqishengchandemubiaomaijin。zuihouyigefangmianshirenlixiaolv
，yishizhongzhongzhizhong。dangqian，zhizaoyezaizhaopinshuliangongrenfangmiancunzaizhuduotiaozhan。zhizaoyebixujinkenengdijianshaorenweijieru，kecaiqudefangshibaokuo：推廣自主製造模式，應用先進機器人技術，部署具備實時感知能力、能快速響應操作環境與製造需求變化的自動化解決方案（見圖3）。

圖3.資源感知型製造。

未來數字化工廠

ADI公司對未來數字化工廠的願景，聚焦於連接、控製和解讀這三大核心支柱。連接戰略旨在通過提升製造業生產效率、可ke擴kuo展zhan性xing和he靈ling活huo性xing，同tong時shi降jiang低di碳tan排pai放fang

，來lai達da成cheng未wei來lai工gong廠chang的de發fa展zhan藍lan圖tu。確que保bao所suo有you製zhi造zao資zi產chan和he機ji器qi連lian接jie到dao統tong一yi網wang絡luo，實shi現xian製zhi造zao數shu據ju的de透tou明ming訪fang問wen，並bing利li用yong這zhe些xie數shu據ju推tui動dong整zheng個ge製zhi造zao場chang所suo的de工gong藝yi持chi續xu改gai進jin。製zhi造zao環huan境jing須xu借jie助zhu有you線xian和he無wu線xian混hun合he網wang絡luo，實shi現xian從cong邊bian緣yuan到dao雲yun端duan的de實shi時shi無wu縫feng連lian接jie。對dui於yu有you線xian控kong製zhi連lian接jie，千qian兆zhao位wei工gong業ye以yi太tai網wang正zheng被bei部bu署shu用yong於yu工gong廠chang網wang絡luo以yi提ti供gong更geng高gao的de帶dai寬kuan，同tong時shi搭da配pei時shi間jian敏min感gan型xing網wang絡luo(TSN)來確保實時流量控製的確定性。對於諸如自主移動機器人(AMR)等移動應用，靈活的專用5G網絡起到補充作用，並且專用5G網絡還可連接難以輕鬆接入有線工業以太網的遠程傳感器和執行器。

dierxiangguanjianzhanlvejujiaoyukongzhilingyu。fensanshizizhukongzhiyituoquanxindemokuaihuazidonghuajiejuefangan，dailaigenggaodelinghuoxing，jinengsuoduanshezhiheyanzhengshijian，younengzhichiriyizengchangdexinchanpinkucundanwei(SKU)。從傳統生產線的集中式可編程邏輯控製器(PLC)轉向分散式PLC控(kong)製(zhi)

，先(xian)進(jin)的(de)邊(bian)緣(yuan)計(ji)算(suan)將(jiang)被(bei)直(zhi)接(jie)集(ji)成(cheng)到(dao)機(ji)器(qi)之(zhi)中(zhong)。基(ji)於(yu)邊(bian)緣(yuan)的(de)自(zi)主(zhu)控(kong)製(zhi)讓(rang)生(sheng)產(chan)線(xian)更(geng)具(ju)可(ke)重(zhong)構(gou)性(xing)
，顯(xian)著(zhu)提(ti)升(sheng)製(zhi)造(zao)靈(ling)活(huo)性(xing)。每(mei)一(yi)台(tai)機(ji)器(qi)都(dou)成(cheng)為(wei)一(yi)個(ge)完(wan)整(zheng)獨(du)立(li)的(de)模(mo)塊(kuai)化(hua)製(zhi)造(zao)單(dan)元(yuan)，可(ke)在(zai)極(ji)少(shao)人(ren)為(wei)介(jie)入(ru)的(de)情(qing)況(kuang)下(xia)，輕(qing)鬆(song)完(wan)成(cheng)配(pei)置(zhi)與(yu)重(zhong)新(xin)部(bu)署(shu)。通(tong)過(guo)部(bu)署(shu)更(geng)多(duo)靈(ling)活(huo)、模塊化的製造解決方案，並由分散式自主控製予以支持，我們能夠更好地實現未來數字化工廠的目標。

zuihouyixiangzhanlvejujiaoyujiedu。jieduzhanlvezhizaijiangshengchanshujuzhuanhuaweikefuzhushijiandedongchaxinxi，congerzhulishixianweilaigongchangdegexiangmubiao。jugusuan
，zhizaoyemeinianchanshengdeshujuliangyueda1812 PB（拍字節）。2 解(jie)讀(du)戰(zhan)略(lve)將(jiang)運(yun)用(yong)人(ren)工(gong)智(zhi)能(neng)技(ji)術(shu)來(lai)處(chu)理(li)這(zhe)些(xie)海(hai)量(liang)製(zhi)造(zao)數(shu)據(ju)，以(yi)提(ti)升(sheng)生(sheng)產(chan)效(xiao)率(lv)。解(jie)讀(du)戰(zhan)略(lve)的(de)關(guan)鍵(jian)在(zai)於(yu)在(zai)數(shu)據(ju)產(chan)生(sheng)的(de)邊(bian)緣(yuan)側(ce)部(bu)署(shu)人(ren)工(gong)智(zhi)能(neng)。邊(bian)緣(yuan)人(ren)工(gong)智(zhi)能(neng)將(jiang)通(tong)過(guo)主(zhu)動(dong)決(jue)策(ce)
，結(jie)合(he)傳(chuan)感(gan)器(qi)融(rong)合(he)（包含工業視覺、溫度
、壓力/力
、測斜儀
、位置、振動
、濕度等測量方式），shixianzhizaoyedezizhuyouhua。bianyuanrengongzhinengjiangtongguozidongzhixingchangguirenwu
，jianshaoduishulianlaodonglideyilai，bingyijinkenenggaodeliangpinlvshixiangengjugexinghuahefuzaxingdezhizao。guanjianyingyongbaokuoyindaoqudong（移動機器人）、缺陷或異常檢測（機器健康狀況）、持續的工藝改進
、模式識別（質量控製），最終還將融入自動化控製循環
，成為其中重要一環。

圖4.實現未來數字化工廠的幾點關鍵要求。

結論

製造業正在經曆一場變革，朝著更智能
、更互聯、以(yi)軟(ruan)件(jian)定(ding)義(yi)為(wei)主(zhu)的(de)方(fang)向(xiang)發(fa)展(zhan)。實(shi)時(shi)無(wu)縫(feng)的(de)邊(bian)緣(yuan)到(dao)雲(yun)端(duan)連(lian)接(jie)，將(jiang)實(shi)現(xian)對(dui)新(xin)型(xing)製(zhi)造(zao)數(shu)據(ju)集(ji)的(de)透(tou)明(ming)化(hua)訪(fang)問(wen)。分(fen)散(san)式(shi)控(kong)製(zhi)借(jie)助(zhu)邊(bian)緣(yuan)計(ji)算(suan)，將(jiang)控(kong)製(zhi)功(gong)能(neng)從(cong)可(ke)編(bian)程(cheng)邏(luo)輯(ji)控(kong)製(zhi)器(qi)(PLC)遷移至機器本身。傳感器融合技術的應用提升了機器的設備綜合效率(OEE)，並bing產chan生sheng豐feng富fu的de數shu據ju集ji，為wei人ren工gong智zhi能neng模mo型xing的de訓xun練lian與yu部bu署shu提ti供gong支zhi撐cheng。邊bian緣yuan人ren工gong智zhi能neng將jiang使shi自zi動dong化hua機ji器qi完wan全quan實shi現xian自zi主zhu化hua。這zhe些xie新xin技ji術shu的de融rong合he勢shi必bi將jiang徹che底di改gai變bian未wei來lai的de數shu字zi化hua工gong廠chang，在zai顯xian著zhu降jiang低di能neng源yuan消xiao耗hao和he材cai料liao浪lang費fei的de同tong時shi，提ti高gao製zhi造zao業ye的de生sheng產chan效xiao率lv
、linghuoxinghekekuozhanxing。duiyuzhizaoshangeryan，chenggongdeguanjianzaiyuruheyushengtaixitongneideqitagongsizhankaihezuo，yinweifengfuduoyangdejingyanhenengliduiyujiasushixianweilaishuzihuagongchangdeyuanjingzhiguanzhongyao。ruxujinyibulejieADI針對未來數字化工廠的可持續自動化解決方案
，請訪問

analog.com/industrialautomation。

參考文獻

1 Victor Reyes

、Heather Ashton和Chad Moutray

，“Creating Pathways for Tomorrow’s Workforce Today:Beyond Reskilling in Manufacturing”，Deloitte Insights，美國製造業研究所，2021年5月。

2 “Deloitte Survey on AI Adoption in Manufacturing”
，Deloitte，2020年。