誰怕千禧蟲?──台灣產業迎戰Y2K

:::

1999 / 4月

文‧李光真 圖‧卜華志



去年底,聯合國召開緊急會議,全球共有一百多國參加,並考慮成立「Y2K國際危機小組」。有人形容這是一場世界大戰,只是當前的大敵不是尋常血肉之軀,而是科技史上最大的 烏龍產物──電腦千禧蟲。令人心驚的是,隨著時間飛逝,人類要想全身而退,似乎已成奢望。

同時,千禧蟲之戰,也將連帶牽動各國產業競爭力的消長。有些國家已成國際投資人抽腿避離的高風險戰區,有些國家則可望在安度蟲蟲危機後,享受訂單蜂擁而至的繁榮果實。

台灣產業,將會落敗,還是歡慶豐收?

「希望到了二千年元旦凌晨,各位看到的燦爛煙火是慶祝的,而不是哪個工廠爆炸……,」在一場替產業資訊人員舉辦的Y2K(Year 2000,即「西元二千年資訊年序危 機」)演講中,主講人台灣IBM公司大中華地區公元二千年綜合策進處協理林獻仁做了以上結論後,台下二百多位聽眾悄然無聲,人人一臉凝重。

有那麼嚴重嗎?

近來,俗稱「千禧蟲」的Y2K議題逐漸在報章上加溫,然而,面對工業局及各產業公會不斷的訪談抽樣,不少企業老闆都有個共同的疑惑:「真有那麼嚴重嗎?」

根據世界著名經濟機構 Gartner Group 對全球八十多個主要經濟國的評估,Y2K的因應進度可分為四級(表一)。A級國家儘管進度領先,但估計仍有百分之十五 的企業會在千禧蟲攻擊下遭受衝擊;這個比例在B級驟升一倍,達到三分之一,包括台、韓、星等小龍都屬於這一級。C級有一半的企業將慘遭重擊;至於第四級,受衝擊比例更高達 三分之二。

這個預測夠嚇人吧!但林獻仁還認為這種評估太樂觀,並沒有把企業倒閉的「骨牌效應」納入考慮。近兩年來四處奔走,協助企業處理Y2K問題的資策會系統工程處副處 長王文溥也 強調,Y2K危機,危害之烈,超過金融風暴,而且是「知道越多越害怕、做的越多越沒有把握!」 到底,千禧蟲是什麼怪物,能讓資訊老手們都聞之色變? 回到一九○○……

探究原因,Y2K其實全無陰謀密辛,只能以「烏龍」視之:原來在六○年代電腦發展初期,每一位元的記憶體都昂貴如金,為了節省記憶體空間,程式設計師只用兩個欄位來代表年 份,而將代表世紀的「一九」省略。沒想到這種當時看來聰明合理的設計,卻在時序進入西元兩千年時衍生出大問題。

「過了今年,電腦裡的年份記錄為○○,請問它如何判斷這是一九○○,還是二○○○?」凌群電腦公司市場處總工程師陳建國解釋,加上有些程式設計師習慣用「九九」或「○○」 代表「最大值」、「程式結束」等特殊含意,而西元二千年又是逢百整數年中每四個百年才出現一次的潤年,使得電腦年序問題更形棘手。

電腦的本質是計算機,而許多計算、比對及排序的基準都和日期或年份有關。於是近幾年全球陸續傳出一些怪現象,譬如信用卡發卡銀行無法發出使用期限跨越西元二千年的新卡,航 空公司無法接受二千年後的機票預約,雜誌社訂戶收到「貴訂戶將於一九○一年到期」之類的通知……等,各國政府及企業才發覺事態嚴重。

對產業界來說,一個程式設計上的小失誤,竟可能導致龐大的生產線當機,似乎難以置信。然而從兩年前就開始大聲疾呼、希望喚起國人注意的英商聯合利華公司桃園廠總工程師康之 政,就經歷過這種夢魘似的打擊。 「老闆,工廠停掉了……」

康之政指出,他的部門工程師曾經對廠裡所有的機器設備做過單機測試,從來沒有發現異常,直到接到英國總公司的文件,指出同類型設備已經偵測出千禧蟲時,他還覺得不服氣。於 是康之政和生產部門講好,在某天下班前將生產部控制室的電腦主機設定為一九九九年十二月三十一日晚上,等著第二天早上去看結果。

結果第二天清晨,當他進入控制室時,發現現場工程師們都神情古怪,原來整座工廠猶如死城,所有機器停擺,電腦監視畫面不動,滑鼠也失靈了!這可不是隨便按個「reset」 鍵重新開機、再輸入正確日期就可以解決的,公司的損失慘重,康之政則成了「罪人」。

當機,停工,萬事休矣。不過另一種更令人不寒而慄的,則是電腦表面上運作如常,其實已悄悄走樣、失真,或是資料檔案莫名其妙被毀損。

石化原料PTA產量佔全球八分之一的中美和石化公司,在進行生產線的Y2K整合測試時,就赫然發現螢幕上的所有圖形、曲線都呈現扭曲異常,原來是scale(平均度)亂掉 了,「幸好有測!」中美和Y2K專案負責人、資訊部經理林明宗當時第一個念頭就是慶幸。

康之政也指出,聯合利華生產旁氏化妝品的工廠,在測試後也發現產品有過酸、過鹼、濃度過高或過低的各種變質現象。換是一家對Y2K不夠重視的化妝品廠商,結果如何,恐怕會 令愛美的女士們思之色變吧?

當心你的大哥大!

在此,惠普科技公司Y2K負責人、顧客服務處軟體支援服務部協理吳增峰特別指出,一般企業內的千禧蟲問題可以分為兩部份。一是資訊系統部份(MIS),譬如財務、人事、訂 單處理等,由於這些資訊在電腦上運作,目標顯著,任何人都知道裡面可能有千禧蟲藏身,必須以軟、硬體升級的方式來解決。

然而,除了電腦外,其他外表和功能五花八門的各式機器設備,也可能暗藏危機,而且更為「無解」。

「工具機、空壓機、鍋爐、儀表、電梯、影印機、傳真機、打卡機、收銀機,甚至大哥大和BB call……,只要想得到的科技產品,裡面都可能有嵌入式晶片(微處理器),」 吳增峰解釋,這些晶片只要內含日期設定、或是它的運作和時間有關,跨年時就可能「中毒」、爆發千禧蟲病。

主導國內中心──衛星工廠體系的中衛發展中心專案經理姜述尚則指出,過去十多年來,全球平均每年生產七十億顆晶片,估計其中約百分之二和日期或年序相關,會發生Y2K問 題。

百分之二,聽起來比例不高,但十幾年累積下來,就意味著有千禧蟲藏身的晶片高達數億顆;若加上各工廠為特殊用途而自己寫程式、自己燒錄的特殊晶片,數量更是可觀。一般推估 一座自動化程度較高的工廠中,約有百分之十至十五的機器設備可能遭千禧蟲污染,嚴重性不可低估。

尤其在自動化程度高的工廠中,整個生產過程都用電腦來監控,譬如食品工廠的電腦依據訂單,自動排定產品配方與生產時程,庫房則依據時程自動提取原料;原料何時進入生產線、 數量多少、由誰負責,完全由電腦掌控。這種「程監控系統」若出問題,就算線上的單台機器沒有問題,生產線也會大亂。

對廠商來說,生產線的千禧蟲,遠比辦公室裡的資訊系統來得棘手,因為多數嵌入式晶片根本無法讀取或更動裡面的內容。出了問題,別說廠商,就連資訊業者都無法進行修改。一般 的作法是發函給原廠供應商,請他確認這型設備是否有Y2K問題?如果有,是否有合格的晶片可以升級?再不行,就只好撒大錢整台汰換了。

別輕信「廠商保證書」

找原廠商負責,困難度高不高?

「應該還好啦,」第一大本土藥廠、上市公司永信藥品的發言人洪士立經理表示,永信使用的特用化學儀器如膠囊充填機等,都是向義大利或日本等國際知名大廠買的。大廠信譽好,又是長期往來的,不會不負責任。然而,另外一家企業的資訊主管卻不如此樂觀。他指出,有些大廠「獅子大開口」,換裝軟體要收取高額服務費,還一直暗示乾脆買新型機種才有保障,讓他左右為難。

儘管不少國內外設備廠商願意提供其產品「符合Y2K標準」的確認書,資策會王文溥副處長仍然提醒業者:「越是一下子就拍胸脯保證的,越要質疑他的可信度。反倒那些支支吾吾、考慮好久,最後才發出確認書的,可信度還高一點。」曾經被國內設備廠商「騙」過的康之政也再三警告,「不要輕信廠商的確認函,還是要自己多方測試才行。」

此外,不是每座儀器設備都有「主人」,有些設備廠商早就倒閉、不知去向,成為「孤兒儀器」,業者只好自己進行測試。要注意的是,「不能手癢亂測」,中美和的林明宗警告,若是沒有先做好資料備份就貿然進行測試,結果將儀器中原先設定的參數洗掉,那可就「回不去了」。

林明宗舉例,譬如運用廣泛、在各種設備中都可使用的PLC(可程式邏輯控制器),是種可做各種數字運算比對的儀器。廠商通常會根據不同的用途需要,在裡面設定起初值(各種參數)。譬如鍋爐中的PLC已設定正常的溫度範圍,一旦鍋爐過熱就會自動降溫;萬一測試前後儀器失靈、導致溫度過熱,小則鍋爐毀損,大則引發爆炸,後果不堪設想。

小企業,安啦?

截至目前,能就Y2K議題侃侃而談的,幾乎清一色是各產業中具有龍頭地位的中大型企業。大企業有錢、有人、有資訊,尤其國際化程度深、或是有國際大廠做技術母廠的,心態上 更是篤定許多。像中美和的大股東美商亞美和(AMOCO),不僅要求逐月審核Y2K進度報告,也派人來台灣輔導。

此外,做為台灣出口產業主力的電腦及半導體廠商,多半都是替大型跨國企業作OEM或ODM,若做不好,馬上就會被抽單;而國外客戶為了確保貨源,還不時跨海來台緊迫盯人, 因此這些企業都是產業主管機關眼中的Y2K模範生,應可順利過關。

然而,「這樣的企業畢竟是少數,不會超過百分之五,」中衛中心姜述尚指出,台灣製造業家數高達八萬多家,大部份是螞蟻雄兵、因陋就簡式的中小型企業。以往中小企業一直是台 灣經濟奇蹟的主力尖兵,在千禧蟲危機中卻可能淪為高危險群。

直至今天,距離二千年只有二百多天了,多數中小企業依然對千禧蟲無甚概念,甚至嗤之以鼻。今年春節前才在全省「走透透」兩圈半,進行演講輔導的林獻仁也估計,不管千禧 蟲議 題吵得多麼火熱,估計約有百分之十幾至二十的企業,就是鐵了心腸、不動如山。

一位名列青創會「工商張老師」、經常為中小企業進行輔導的資訊服務業者,對中小企業的心態倒也「見怪不怪」。

這位業者認為,大企業自動化程度高,什麼都用電腦連線來控制,才會有Y2K問題,至於一般中小企業,工廠裡很少有整合性的程監控系統,頂多就是幾座工具機必須汰換,「換就 換嘛,反正只是幾十萬的事情。」

他甚至認為,就算到了西元二千年元旦當天突然當機,也不是什麼世界末日:「中小企業從訂料到出貨、收款,兩三個月就完了,誰會用到跨年度的長時序?再說,中小企業的生產 線、裝配線走走停停是常事,有什麼奇怪?他們的財務簡單,就那一兩個月用手算一下,又有什麼關係?等所有作業都脫離一九九九、變成○○減○○的時候,不就正常了嘛!」

「安啦,大企業自動化程度高才要擔心,中小企業何必去管它?這些都是外國少數國家、少數幾個愛作秀團體的危言聳聽。他們還不是想賣機器!」他語帶不屑地說。 管理的危機

自動化程度低,「理論上」受衝擊的層面的確比較小,但另一方面,中小企業的資源少,手頭緊,對衝擊忍受的能力也比較差。再說,「到底一座中小型工廠裡面會有多少機器出問 題,沒做過清倉、測試,誰也不敢說,」姜述尚指出。

以成衣加工廠來說,製衣用繪圖機、自動裁切機、自動打版機,已普遍發現Y2K問題,可是有 多少業者知道這些機器暗藏危機?如果拖到當機了才手忙腳亂要買新機器,財務上調度 得過來嗎?客戶願意等嗎?

對Y2K的認知不足,也使多數中小企業老闆認為,「既然是電腦出錯,讓資訊部門負責就好了嘛」,卻忽略了最重要的、生產部門的程監控問題,結果問題還是無法解決。

「我們資訊部門是懂電腦沒錯,可是我們不懂生產啊!」一家股票上市的化學公司資訊部副理沮喪地指出,他雖是公司的Y2K專案負責人,但沒人、沒權,什麼事也推動不了。

「我只是掛名的!」這位副理一再強調:「我對生產部門的程監控根本一竅不通,完全外行!」 如果真要測試,生產線必須停機,但他一個小小資訊副理,怎麼可能要求員工數百人的 生產部門停工配合?

荒謬的是,對照這位資訊主管的惶恐,該公司生產部門經理卻老神在在地表示,「我們生產製程應該沒影響,因為公司資訊部門很早就開始注意這件事了, 他們都有在追蹤,沒聽說有 什麼問題……」

另一家股票上櫃的機械公司資訊部門主管則表示,他們的資訊系統已修正完畢,而他雖然對工廠的生產作業不熟,但「有信心」不會有問題:「我們的機器都是沖床、洗床,或是塗裝 機。只有和溫度、濕度有關的,沒有和日期有關的!」他堅稱根本不需要測試,「我們沒事去輸入日期幹什麼?就算機器到時候日期錯誤又有什麼關係?產品上面又不打日期!」

然而,「工廠的儀器設備和程監控,與資訊系統根本是兩個完全不同的領域,」惠普吳增峰表示,儀器內的晶片是否和日期相關,除了將晶片內的程式碼解讀出來外,根本不可能單憑 外觀判斷。 董事長們請出列!

正因為看到許多企業內資訊人員的窘境和不足,吳增峰強調,Y2K「絕不光是電腦技術的問題,更是整個企業經營管理的問題。」唯有企業主用統合式、高層次角度去看待Y2K議 題,才能發揮戰力,殲滅千禧蟲。

以在石化業中游產業的Y2K模範生中美和來說,他們的測試十分完整,先清倉(找出所有可能受污染的設備)、修改後,再依序進行單元測試、整合測試,及以正常開機狀態運轉的 「進水測試」。

「這些測試必須和生產部門協調,利用五個廠分別年度歲修的時候進行,以免干擾到正常的出貨,」中美和經理林明宗表示。尤其第一廠進行「進水測試」時,由於是全球創舉,整整 停工十五天,比預定進度多耗費了一、兩天,這也必須由管理階層慎重評估、全力支持才行。

事實上,許多Y2K成功案例,背後都透露出管理者的遠見。

剛就任華擎機械公司總經理的前裕隆汽車發言人兼Y2K負責人林茂寅副總經理指出,裕隆早在九七年中就著手因應Y2K;而不同於多數企業只以消極抓蟲為目標,裕隆則是趁此機 會,徹底改善電腦系統,「整個企業二十二個電腦系統中,除了一個做局部修正外,其他全部重新開發,包括組裝線控制(ALC)、原物料計畫(MRP)、研發設計(PDM) 等,」林茂寅指出,目前已有十四個系統完成上線,還有八個正在加緊努力中。 兼善天下以求自保

進一步從技術和經營的層次跳脫,有些業者更開始從「營運──企業如何存活」的角度去考量 Y2K議題。

「企業可以關起門來,把自己家裡的千禧蟲都抓光光;可是如果外面的環境毀壞了,光你獨善其身也活不下去,」林獻仁指出。

放寬視野,企業所處國家的水、電、交通、關務等基礎建設合不合格?上游的原物料供應商、各零件衛星工廠及協力廠,還有下游的重要客戶、經銷體系、物流體系……,每一個環節都可以安然過渡到西元二千年嗎?如果不能,對企業會造成哪些衝擊?這些都是需納入考慮的變數。

「業務是企業的根本,若是通路垮了,東西再好也賣不動,」林茂寅指出,裕隆早有這種危機意識。為了確保通路無礙,裕隆不僅免費為全省一百三十多個經銷商舉辦教育訓練,免費提供升級軟體,務必「盯」到每家經銷商都確實做好Y2K因應措施。

經銷商只賣裕隆的車子,因此裕隆可以這樣嚴格要求。然而面對全省林林總總、大大小小的一百多家協力廠商,包括生產烤漆顏料、椅套、甚至做螺絲的,裕隆頗有無處使力之感。

「全省有十三家車廠,協力廠不是只賣我們裕隆一家,」林茂寅指出,面臨千禧蟲大敵,有些車廠想這樣改、有些想那樣改,公公婆婆太多的結果,協力廠商也無所適從。幸好今年春節過後,全國車輛公會在工業局的要求下「站」出來協調,由五大車廠各自負責「認養」部份協力廠,裕隆被分配到六十一家。雖然時間緊、工作重、壓力大,但總比先前的無從下手要好。

林茂寅認為,在企業自顧不暇時,主管產業的經濟部工業局適時介入,幫助頗大。事實上,自從去年工業局出面整合,並且在工研院、資策會、中衛中心等素來和產業界往來密切的單位聯手下,已逐漸在業界醞釀出一股合作氛圍。許多做的相當成功的企業,如筆記型電腦大廠英業達等,都陸續設立Y2K戰情室、Y2K網站,將自己的經驗公開,讓各界分享。

滅蟲總動員

最近工業局更開始透過同業公會的力量,加強產業間的橫向聯繫。

「同一產業可能使用的機器設備往往大同小異,如果每家企業都能將自己針對某廠牌、某機型設備做測試的結果彙整起來,中小企業就可以不用自己摸索,對他們幫助很大,」中衛中 心姜述尚指出。

林獻仁則強調,台灣產業彼此的依存度極高,環環相扣,是不折不扣的「生命共同體」。「產業界對什麼議題都可以互打互搶,只有Y2K不是用來競爭的,必須合作才能解決,」林 獻仁強調。

不過,考量整體產業供應鏈,「海外台商會是最脆弱的一環,」中國生產力中心董事長許介圭指出。

一方面台商雲集的大陸及東南亞地區,正是國際機構  GartnerGroup所評定的高危險區,水電能源等供應的穩定度堪虞;加上當地資訊封閉,台商多屬勞力密集、組裝 型的傳統產業,半自動化老舊設備很多,要毫髮無傷很難。而大陸台商若受傷慘重,台灣每年對大陸的二百多億美元出超能否保住?對台灣產業又會產生什麼衝擊?目前還難以評 估。 是危機,還是商機?

以國際眼光來看,「今年下半年開始,外商就會陸續將訂單從高風險國家、高風險企業中抽出,移轉到他們認為較安全的地方去,」許介圭指出,Y2K即將牽動區域產業競爭力的重 新洗牌。 「台灣做的不錯,有希望力拚亞太第一!」林獻仁指出,台灣的策略正確,政府先盯好基礎建設,再從大型企業開始督促,直到現在的各類組織、各型企業全面動員。

雖然不久前在馬尼拉召開的全球Y2K高峰會,曾經將台灣和德、日並列,點名台灣製造業進度落後,但林獻仁在仔細研究美國的研究報告後,覺得台灣並不遜色。如今時程雖然緊 迫,但以台灣中小企業一向與政府產業輔導單位往來密切,加上自動化程度較低、因應起來較單純的情況評估,並不悲觀。

相較之下,「新加坡和香港做得不錯,但是他們高度依賴的腹地──馬來西亞和大陸,卻是大有問題,」林獻仁指出,台灣在佔有相對優勢的情況下,不 妨掌握契機,爭取國際的訂 單。

Y2K,是危機,還是商機?未來半年的衝刺至關重要。台灣產業,可有信心迎向挑戰?

相關文章

近期文章

EN

Who's Afraid of Y2K? Taiwan Business Takes on the Millennium Bug

Laura Li /photos courtesy of Pu Hua-chih /tr. by Phil Newell


At the end of last year more than 100 countries participated in a special United Nations conference which decided to establish a "Y2K international crisis team." Some people say "it's a war out there." But there's no enemy out for blood. Rather we're up against an accident of technological history-Y2K. Most worrisome is that, for most people, it is already too late to stay out of the line of fire.

The Y2K war will affect the productivity of industries in countries around the world. Some countries are "high-risk areas," from which investors are fleeing for safety. Others can look forward to getting through the crisis rather smoothly, or even enjoying an upsurge in orders.

Will Taiwan industry be one of the losers, or one of the beneficiaries?

"I hope that in the early hours of the morning on New Year's Day of the year 2000, the sparks in the air will be from fireworks, and not from a factory exploding." This was the concluding comment made by Austin Lin, manager of the Year 2000 group at IBM Taiwan, during a Y2K seminar held for information personnel from the commercial world. His audience appeared to be in a state of shock from his speech.

Over the past half year or so, Y2K has become an increasingly hot subject in the media. Nevertheless, sample surveys conducted by the Industrial Development Bureau (a division of the Ministry of Economic Affairs) and by industrial associations reveal that many entrepreneurs are sceptical: "Is it really that serious?"

The Gartner Group, a well-known international economic institution, has released an assessment of how well more than 60 of the economically most important countries are handling Y2K. They divide these countries into four categories (see Table 1). Those in Category 1 are doing the best in coping with Y2K, yet an estimated 15% of their enterprises will suffer serious operational impact from Y2K. The proportion of affected firms doubles for Category 2 countries, to one in three. Taiwan, Korea, Singa-pore, and Hong Kong are in this category. In Category 3 countries, half of enterprises will be seriously affected, and in Category 4 countries the figure will be two-thirds.

Shocking, isn't it? However, Austin Lin believes that even this is too optimistic, because it does not include the "domino effect" on other enterprises. W.P. Wang, deputy director of the Systems Engineering Division at the Institute for Information Industry, who has been running himself ragged the last two years to assist industry with its Y2K problems, emphasizes that the Y2K crisis is even more serious than the recent financial meltdown. He adds, "The more you know the more frightening it is, and the more you do the more out of control it seems!"

What is it about Y2K that makes information technology veterans blanch at its mere mention? Return to 1900

There is no dark plot underlying Y2K. Rather, it is an accidental byproduct of technological history. In the early days of computer development in the 1960s, every bit of memory was precious. To save space, engineers abbreviated years by using only the last two digits. Little did anyone expect that a design which at that time appeared to be reasonable would, with the coming of the year 2000, become a major problem.

Jack Chen, chief engineer in the marketing division at Syscom Computer Engineering Company, explains: "After this year, the year indicator in computers will turn to 00. I ask you, how is the computer to tell whether this means 1900 or 2000?" The problem is complicated by the fact that some computer programmers routinely used 99 or 00 to indicate special meanings like "largest value" or "terminate program." And don't even mention the fact that 2000 is one of the rare century years which is also a leap year, making the computer trans-millennial dating problem even thornier.

Computers are fundamentally calculators. Many of their calculations and procedures are connected to the date. Because of this connection, strange things have started to happen in recent years. Credit-card-issuing banks found that they could not issue new cards which could be used beyond 1999. Airlines were unable to accept reservations for 2000. And periodical subscribers received bizarre notices like "your subscription will expire in 1901." Governments and enterprises began to take notice.

For business, the possibility that a small computer design error can lead to a major breakdown of the production line might seem a little hard to believe. Nevertheless, that is exactly the nightmare that happened to Kang Chih-cheng, an engineer at the Taoyuan plant of the multi-national firm Unilever, who has for the last two years been trying to draw attention to this problem in Taiwan. "Sorry boss, but the factory is down."

Kang says that his department's engineers did complete machine-by-machine checks of all their equipment, and discovered nothing unusual. But when he received notification from the Anglo-Dutch parent company that their tests had found Y2K problems in similar equipment, he began to worry. After discussions with the production department, it was agreed that on a certain day, just before the end of the workday, the date in the main computer in the production department control room would be changed to December 31, 1999, and then they would see the next morning what had happened.

The next morning, when Kang walked into the control room, all the engineers looked as if something bad had happened. Not surprising, considering that the whole factory was about as lively as a morgue. Not a mechanical creature was stirring, not even a mouse. This was moreover not a problem that could be resolved simply by pressing the reset button and entering a new date into the computer. The factory suffered serious losses and Kang was blamed.

It's bad enough when everything grinds to a halt. But it's even scarier when everything appears to be running normally on the surface, but things are quietly going awry. Perhaps files are mysteriously disappearing. . . .

When conducting its own overall Y2K tests, China American Petroleum Company, which accounts for one-eighth of the world's production of the material PTA, was shocked to discover that all the diagrams and lines on the computer screens appeared distorted. It turns out that the scale got completely screwed up by Y2K. "It's a good thing we checked ahead of time!" exclaims Bill Lin, CAP's information service manager and Y2K task force director.

Kang Chih-cheng adds that when the factory that makes Ponds cosmetics products for Unilever did Y2K tests, they experienced problems, such as excessive acidity or alkilinity, or uneven consistency, with many products. Women who are concerned about their looks might go pale at the very thought of what might happen if their beauty products come from a factory that doesn't take Y2K seriously.

Mobile mayhem

Wu Tseng-feng, business manager of Software Services and Support at Hewlett-Packard Taiwan, divides the Y2K problem into two parts. The first involves data systems, such as financial affairs, personnel, and customer orders. Because the data is stored in computers, the target is obvious. Everyone knows that Y2K could be lurking, so they can protect themselves with software and hardware upgrades.

But all kinds of equipment and machinery other than computers might also be harboring Y2K.

Wu states: "You name it. Any product that uses advanced technology-machine tools, air compressors, kilns, instruments, elevators, printers, faxes, time card machines, cash registers, and even mobile phones and pagers-could have an embedded microprocessor." If these chips have internal dates, or if their functioning is connected to time, they might come down with a bad case of Y2K at the turn of the century.

An estimated 2% of the chips produced in the past decade are connected in some way to the date or the year, and may experience Y2K problems. That seems like a pretty small number. But, says Chiang Shu-shang of the Corporate Synergy Development Center, in the 1990s an average of more than seven billion chips a year were produced worldwide. Thus several hundred million chips could have Y2K problems. The number is even larger if you add on chips made by factories themselves for special in-house purposes. It is estimated that in a factory with a fairly high degree of automation, approximately 10-15% of the equipment could be affected by Y2K. The threat is not to be underestimated.

In highly automated factories, the entire production process is monitored and controlled by computers. For example, in some food processing factories, computers automatically produce product formulas and production schedules based on incoming orders. The warehouse provides raw materials based on the schedules. If the computer deciding the schedules and deliveries experiences problems, even if not a single machine on the production line has a Y2K bug, the line could still end up in chaos.

For businesses, Y2K poses a much bigger threat on the production line that it does in the office. This is because most embedded chips cannot be read or altered. If there's a problem, forget about the factory itself, even computer geeks would be unable to help.

Most factories are dealing with Y2K by sending queries to their suppliers: "Does the machine you sold us face a Y2K problem? If so, is there a replacement chip to upgrade the unit?" If there is not, companies may have to replace the whole machine.

Don't believe it!

Will original manufacturers take responsibility?

"It shouldn't be a big problem for us," says Hung Shih-li, manager of Taiwan's largest pharmaceuticals firm, Yung Shin. Their equipment has been purchased from large well-known international makers in places like Italy and Japan. These manufacturers have good reputations, and value long-term relations, so they are unlikely to evade their responsibilities.

However, the information manager at another company is not so optimistic. His experience has been that many factories are demanding high service fees to take back their software, and continually hint that the only real guarantee is to replace the whole machine. He is stuck between a rock and a hard place.

Many domestic and foreign equipment manufacturers are perfectly willing to supply forms "confirming" that their product is "Y2K compliant." But the Institute for Information Industry's W.P. Wang advises: "The faster they are to put their hands on their hearts and swear up and down, the more dubious their promises are. On the other hand, you can probably put a little more trust in those factories that consider for a long time before finally issuing a confirmation." Kang Chih-cheng, who has been "deceived" by domestic manufacturers before, never tires of warning: "Don't readily believe the guarantees given by manufacturers. It's still best to test things yourself."

Of course, not every machine has a "parent." Some are "orphans" whose manufacturers have gone out of business. Businesses have to test these themselves.

Regardless of whether or not there is a manufacturer around, it is vital to remember this: Don't test haphazardly. Bill Lin cautions, "Never test without proper preparation." If data has not been properly backed up first and testing goes ahead, the originally installed parameters in the chips could be wiped out, and at that point there would be no going back.

Lin offers a case in point. The programmable logic controller (PLC) is widely used in all kinds of equipment. It can check and correct all kinds of numerical operations. Depending upon their needs, factories will set basic values (parameters) into a PLC. For example, a PLC in a kiln could be set for the range of normal temperatures. If the kiln gets too hot, the PLC will automatically lower the temperature. If the chip were to be damaged by testing and the temperature ran wild, the kiln would certainly be ruined and there could even be a lethal explosion.

Small and out of the way?

So far, industry leaders have been the main players discussing Y2K in detail. These large enterprises have money, manpower, and information. The ones in the best position of all are those enterprises with a high degree of internationalization or who rely on international parent companies for their technology. For example, China American Petroleum's major stockholder, the US firm Amoco, not only requires monthly progress reports on Y2K, it has even sent people to Taiwan to consult.

Things likewise look OK for computer and semiconductor makers, Taiwan's leading exporters. Most of these companies do OEM or ODM work for multinational corporations. If they do not do well, they will immediately lose their orders. Moreover, foreign customers, to keep tabs on their sourcing, periodically send people to Taiwan to check on things. From the standpoint of government industrial policy managers, these are "model students" in the Y2K game, and they should be able to get through the crisis smoothly.

Nevertheless, as Chiang Shu-shang says, "these types of firms are a small minority, no more than 5%." Taiwan has more than 80,000 manufacturers, most of which are low-tech SMEs (small and medium-size enterprises). Formerly the crack troops in Taiwan's economic miracle, in the Y2K crisis they may be the highest risk group.

Although there are only about 200 days left before 2000, most SMEs have little idea about Y2K, and some even scoff at it. Austin Lin, who just before the Chinese New Year completed a two-and-half-circuit tour of the island lecturing and consulting on Y2K, estimates that no matter how much attention is devoted in the public discourse to this problem, an estimated 10-20% of enterprises will still have closed minds about it.

On the other hand, one entrepreneur in the information service industry who is a commercial adviser for the ROC Junior Chamber of Commerce and often does gives advice to SMEs, finds the attitude of many SME bosses understandable.

This businessman argues that it is mainly large enterprises that face serious Y2K problems, because they are the only ones with a high degree of automation, using computer networks to control everything. SMEs, meanwhile, rarely have integrated process control systems. At most they may have to replace a few machines. "If they have to replace them, they'll replace them, anyway it's just a few hundred thousand NT dollars."

He argues that, even assuming that on New Year's Day 2000 some machines crash, it wouldn't be the end of the world. "The commercial cycle at an SME, from ordering their materials to shipping their product to collecting the money, takes about two or three months. Who uses trans-millennial production schedules? Anyway, they are used to stop-and-go production. Also, their finances are simple, so if they have to count on their fingers for a month or two, what difference does it make? When all of their operations finally emerge from "99" and become "00," then everything will be back to normal, right?"

"Relax," he advises. "Only big enterprises with a high degree of automation have to worry. Why should we smaller businesses worry about it? This is all something cooked up by a small number of foreign firms who like to make a big show, so they are talking as if the sky is falling. I think they just want to sell more machinery!" he says with disgust. Management crisis

In theory at least, firms with a lower degree of automation are indeed less likely to experience a serious shock from Y2K. But SMEs have limited resources, and thus a correspondingly lower ability to withstand shocks. Says Chiang Shu-shang: "How many machines will go down in a given medium-sized factory? Until testing and inventory have been done, no one can say for sure."

Take for example a ready-to-wear clothing factory. Y2K problems have been detected in many of the machines used in such factories. But how many factory operators know that these machines have hidden dangers? If they wait until a machine crashes before they rush off frantically to buy a new one, will they have enough cash on hand to get through? Will their customers be willing to wait?

Many SME bosses admit knowing little about Y2K, but say it isn't their job anyway: "Since the problem is with the computers, let the information systems people take care of it." Unfortunately this overlooks the important problem of manufacturing process control.

The vice manager of the information department in a listed petrochemicals firm moans: "We information department people know computers all right, but we don't know production!" Although he is in charge of the Y2K problem at his company, he has no staff, and no authority.

"It's nothing but an empty title!" says this vice manager. He repeatedly emphasizes: "I don't understand a thing about the process control in our production department, I'm completely out of the loop!" If testing is really to be done, then the production line will have to be shut down. How can a relatively unimportant vice manager of information systems get the production department, with its hundreds of employees, to cooperate and stop work?

In contrast to this panicked information manager, the production manager at the same company is completely at ease. But the reason is comic: "There shouldn't be any effect on our production process, because our information department people have been looking out for this problem for a long time. They've been following up on it, and I haven't heard of any problems. . . ."

Another information department director, this time at a machinery company, states that their information systems have already been completely taken care of, and although he is not familiar with the production operations at the factory, he is "confident" that there won't be any problems: "Our machines are only concerned about temperature and humidity, and don't have anything to do with the date!" He insists that there's absolutely no need for testing. "If nothing is wrong, what's the point of going and plugging in the date? Even if when the time comes, the date is wrong in the machines, what difference will it make? Anyway we don't stamp the date on our products!"

In fact, there's only one real way to tell whether the chips in the instruments are connected to the date, and that is to read the program codes. Otherwise there's no way to tell just by the appearance.

Bosses, get involved!

Having seen the difficult position that the information personnel in many companies are in, Wu Tseng-feng of Hewlett-Packard emphasizes that Y2K "is not simply a computer technical glitch, but involves the entire management structure of the firm." Only if the top-ranking officials coordinate things can full capability be brought to bear in the war against Y2K.

This is precisely the secret of success at China American Petroleum, a "model student" in the Y2K game.

CAP has done extremely comprehensive testing, from an audit (identifying all the equipment which could in some way be affected) and initial adjustments to systematic tests on each machine (individually and then in sets) to a full-scale "flow test" under normal operational conditions. Bill Lin explains: "We had to coordinate these tests with the production department, using the annual repair period of each of the five factories in turn, in order to avoid disrupting normal output."

Particular attention was focused on the "flow test" at the first factory, because this was the first time it had been done anywhere in the world. The whole factory was down for 15 days, which was a day or two more than had been anticipated. This was only possible because there was careful evaluation and full support from high-level management.

Management foresight has been the moving force behind aggressive strategies to cope with Y2K in other companies as well.

David Lin was formerly Y2K coordinator for Yulon Motors. Now the general manager at a different firm, he notes that Yulon began tackling Y2K as early as a 1997. Moreover, unlike many firms which have simply passively waited for problems to crop up, Yulon seized the opportunity to fully upgrade its computer systems. "Of the 22 computer systems in the company, 21 have been completely overhauled, while the last has been repaired," says Lin. Already 14 of the revised systems are online, and work is being rushed ahead on the other eight.

Survival

Besides the technical and management angles, one can also look at the Y2K problem from the point of view of enterprise survival.

"The company can close its doors and kill all the Y2K bugs inside. But if the environment is a mess, you can't survive even if you are taking the proper steps," says Austin Lin.

For example, Y2K could affect the reliability of water, power, transportation, communications, and customs services. Upstream there are raw materials providers, satellite factories, and subcontractors. Downstream are customers, sales networks, and distribution systems. Can each and every link in the chain safely cross over into the year 2000? If they can't, what will be the impact on business? These are all variables that need to be considered.

"Sales is the foundation of a company. If the channels are clogged, products can't be sold no matter how good they are," says David Lin. Yulon took this threat seriously right from the start. They provided free Y2K training to more than 130 sales outlets across the island, plus gratis software upgrades, and then kibitzed over the shoulders of their outlets to make sure that each one was acting to cope with Y2K.

Yulon could be this strict because its retailers specialize only in its product. But they felt powerless faced with their 100-plus large and small subcontractors, producing such things as paint, seat covers, and screws.

As David Lin explains, "There are 13 auto manufacturers in Taiwan, and subcontractors don't just work for Yulon alone." Each factory wanted the subcontractors to do something different to adjust to Y2K, and with so many chefs spoiling the broth, the subcontractors didn't know where to begin. Fortunately, just after the Chinese New Year, the Industrial Development Bureau (IDB) brokered an agreement within the Transportation Vehicle Manufacturers Association, with five major manufacturers each agreeing to "adopt" specific subcontractors and guide them through Y2K. Yulon was assigned 61 subcontractors. Though time is short and there is much to do, at least things are better than when everyone was paralyzed.

Lin says that since every firm was only looking out for itself, it was very helpful that the IDB, which manages industrial policy, stepped in. In fact, since last year the IDB, working with other institutions which have close ties to business-such as the Industrial Technology Research Institute, the Institute for Information Industry, and the Corporate Synergy Development Center-together have created an atmosphere of cooperation in industry. Many firms which are handling Y2K very successfully, such as the notebook computer manufacturer Inventec, have been setting up Y2K "intelligence branches" or websites and making their own experiences available to others.

Full mobilization

The IDB has also been working through industrial associations to build up horizontal linkages between firms, especially SMEs.

"Firms in the same line mostly use the same equipment. If every firm were to make available the results of tests it has done on specific brands and models of machinery, then SMEs wouldn't have to muddle through blindly on their own. This would be tremendously helpful," says Chiang Shu-shang.

Austin Lin stresses that Taiwan's firms are very interdependent. He concludes: "Businesses can fight all they want about anything under the sun-except for Y2K. They have to cooperate to solve this problem."

Looking, then, at the overall production chain, "the weakest link is Taiwanese-owned firms overseas," according to China Productivity Center chairman Hsu Chieh-kwei.

For one thing, overseas Taiwan firms are concentrated in areas that the Gartner Group considers "high-risk"-mainland China and Southeast Asia. Water and power supplies could be even more dubious after Y2K kicks in. Also, those areas are relatively closed to outside information. Most overseas Taiwan firms in Asia are traditional labor-intensive assembly operations, with only a moderate degree of automation and much outdated equipment. It will be difficult for them to get past 2000 unharmed. If Taiwan firms in the PRC are hard hit, what will become of Taiwan's annual US$20 billion surplus with the mainland? What will the impact be on companies in Taiwan proper? Nobody can say for sure.

Crisis or opportunity?

From a global perspective, says Hsu Chieh-kwei, "Beginning in the second half of this year, foreign buyers will begin withdrawing orders from high risk areas and high risk firms, and switch to safer suppliers." Y2K could spark a complete reshuffle in the comparative advantage game.

"Taiwan is doing OK, and can aspire to be doing better than anyone else in Asia," says Austin Lin. Taiwan has adopted the correct strategy, with the government first focusing on core infrastructure, then large firms, and finally going to full mobilization involving various types of organizations and firms of all sizes.

Recently APEC held a summit in Manila which deemed Taiwan's manufacturing sector to be relatively unprepared. But Austin Lin, having followed the research reports coming out of the US, feels that Taiwan is not putting up a poor showing. Though time is short, there is good reason not to be pessimistic about the situation, considering that Taiwan SMEs have a long-standing and solid relationship with government agencies and that SMEs have a relatively low degree of automation.

In contrast, "though Hong Kong and Singapore are doing very well with Y2K, the hinterlands on which they heavily rely-mainland China and Malaysia-are facing serious problems," avers Lin. Taiwan has the comparative advantage, so it can seize this opportunity to pick up more international orders.

Is Y2K a threat or an opportunity? The next half year will be decisive. Can Taiwan's businesses face the challenge with confidence?

X 使用【台灣光華雜誌】APP!
更快速更方便!