1303032 九、發明說明： 【發明所屬之技術領域】 本發明係有關於一種收集半導體製程系統之非標準資料 (atypical data)的相關系統與方法及執行其方法之計算飢渴讀取 記錄媒體’特別係有關於-種有效操作半導體製程系統之資料收 集系、、充一方法，以及執行此方法之計算機可讀取紀錄媒體，其能 收集可靠資料且不需增加半導體製程系統之操作負載。 【先前技術】 近來在半導體工業，大多數的製程設備都可在線上自動操 作。請參考下列第丨圖，其係繪示傳統半導體製程系統的基本架 構。 如第1圖所示，傳統半導體系統基本上包含製程設備11() 與主機120。 製程没備11 〇包含執行特定半導體製程、量測電源、氣壓或 電源線狀態的量測設備、傳輸晶圓或類似裝置到半導體製程線上 後續程序的傳輸設備。 參考一連接到製程設備110的電腦主機120。主機120係用 以控制製程設備110，使之執行一半導體製程系統中一特定的半 導體製程。主機120連結到一資料庫（未顯示），此資料庫儲存著 製程設備110用以執行製程所需的大量資料。 上述的製程設備110與主機120係根據SEMI設備通訊標準 (SECS)以互相交換資訊。 最近引進一系統，用以收集與分析從各種製程設備傳來的資 訊，以從製造觀點來改進生產力。更特別地，蒐集非製程設備所 提供的任意非標準資料的需求增加了。另一傳統製造系統的架構 TW2024(060622)CRF.doc 6 1303032 請參照第2圖與第3圖。 第固係緣示驾知拓展驗證協定（ΕΑΡ)同步資料型半導體 製程系統的架構示意圖。 如第2圖所示，習知半導體製程系統包含製程設備21〇、主 機220、轉換裝置23〇與分析系統24〇。因為資料同步係經由主 機220來執行，上述半導體製程系統可視為一 ΕΑρ同步資料型系 統。 、 上述分析系統240連結至主機220，且透過主機22〇對製程 設備210的狀態作資料蒐集。由分析系統24〇收集到的資料，根 據其型態可以分成兩類：符合SECS標準的SECS資料（以下稱為，， 標準資料”），與不考慮特定型式而隨意獲得的非SECS資料（以下 稱為”非標準資料，，）。 在此例中，同步過程係包括下列程序··從分析系統要求資料 開始的最短時間内，收集到所有必要的資料；安排所獲得的標準 貧料與非標準資料，使這些標準資料與非標準資料可以被安排在 同一臨時輸送轴（temp〇ral axis)上；以及處理所收集到的資料，使 知當收集到的資料傳送至分析系統時，標準資料與非標準資料可 以被處理成一致的標準訊息格式。 在ΕΑΡ資料同步型半導體製程系統中，主機22〇從製造系 統210收集標準資料，以及從轉換裝置23〇收集非標準資料，再 同步整合這些標準型資料與標準非型資料，然後將同步資料傳送 到分析系統240。轉換裝置230獨立於製造系統21〇與主機22〇 之外。轉換裝置230的功能係為轉換各種方式所收集的非標準資 料，使之成為主機220所能辨識的資料型式，並傳送轉換過的資 料到主機220。 同時’在傳統ΕΑΡ資料同步型半導體製程系統中，同步資 TW2024PA 7 1303032 料與事件以實現資料同步的功能必須由主機220額外提供。結 果’傳統的ΕΑΡ資料同步型半導體製程系統中，主機220的負載 會增加，系統的穩定度也會變差。 更進一步，在傳統ΕΑΡ資料同步型半導體製程系統中，因 所有資料都被收集到主機220，造成環繞主機220的網路負載之 增加。此時所產生的負載可能損害系統的穩定度，亦可能阻礙製 造流程的正常進行。結果，傳統ΕΑΡ資料同步型半導體製程系統 的操作可能停機，甚至可能造成嚴重的損失。 此外，傳統ΕΑΡ資料同步型半導體製程系統會因為非標準 資料的低穩定度而產生問題。而上述非標準資料的低穩定度係由 於非標準資料轉換成主機可辨識的形式'或利用網路傳遞傳遞轉 換的資料、或執行資料或事件同步化時的暫時拖延所造成的。 第3圖削會示為一習知代理資料同步型半導體製程系統 (agent data-Sync type semic〇nduct〇r manufactuHng _㈣的架 構示意圖。 如第3圖所示’傳統代理f料同步型半導體製程系統包含製 程設備⑽、主機32〇、代理端33G與分析系統鳩。因為代理端 330進行資料同步，此半導體製程系統可稱為，，代理資料同步型系 機320 理資料同步型半導體製程系統中，代理端330係為主 料，執供的功能。代理端33G收集標準資料與非標準資 科執仃資料同步化，然後傳送資料到分析系統鳩。 端33^可傳統代理資_步型半導體製㈣統中，代理 缺一揭/ :機32〇的系統負擔，但是主機320的網路負擔依1303032 IX. Description of the Invention: [Technical Field] The present invention relates to a related system and method for collecting atypical data of a semiconductor process system and a computational hunger reading recording medium for performing the method There is a data collection system for efficiently operating a semiconductor process system, a charging method, and a computer readable recording medium for performing the method, which can collect reliable data without increasing the operational load of the semiconductor process system. [Prior Art] Recently, in the semiconductor industry, most process equipment can be operated automatically on the line. Please refer to the following figure, which shows the basic architecture of a conventional semiconductor process system. As shown in FIG. 1, the conventional semiconductor system basically includes a process device 11() and a host computer 120. The process does not include 11 量 measurement devices that perform specific semiconductor processes, measure power, air pressure, or power line conditions, transfer devices that transfer wafers or similar devices to subsequent programs on the semiconductor process line. Reference is made to a computer host 120 connected to the process device 110. Host 120 is used to control process device 110 to perform a particular semiconductor process in a semiconductor process system. Host 120 is coupled to a database (not shown) that stores a large amount of data required by process device 110 to perform the process. The process device 110 and the host 120 described above exchange information with each other according to the SEMI Device Communication Standard (SECS). A system has recently been introduced to collect and analyze information from a variety of process equipment to improve productivity from a manufacturing perspective. More specifically, the need to collect any non-standard data provided by non-process equipment has increased. Architecture of another conventional manufacturing system TW2024 (060622) CRF.doc 6 1303032 Please refer to Figures 2 and 3. The architecture of the first solid system is a schematic diagram of the architecture of the synchronous data-based semiconductor processing system. As shown in Fig. 2, the conventional semiconductor process system includes a process device 21, a host 220, a conversion device 23A, and an analysis system 24A. Since the data synchronization is performed via the host 220, the above semiconductor processing system can be regarded as a 同步 同步 synchronous data type system. The analysis system 240 is coupled to the host 220, and collects data of the state of the process device 210 through the host 22. The data collected by the analysis system 24 can be divided into two categories according to their type: SECS data conforming to the SECS standard (hereinafter referred to as "standard data"), and non-SECS data randomly obtained without considering a specific type (below Called "non-standard data,,". In this case, the synchronization process consists of the following procedures: • Collecting all necessary information from the minimum time required by the analysis system to request the data; arranging the obtained standard poor and non-standard data to make these standard data and non-standard The data can be arranged on the same temp〇ral axis; and the collected data can be processed so that when the collected data is transmitted to the analysis system, the standard data and the non-standard data can be processed to be consistent. Standard message format. In the data synchronization type semiconductor process system, the host 22 collects standard data from the manufacturing system 210, and collects non-standard data from the conversion device 23, and then synchronously integrates the standard type data with the standard non-type data, and then transmits the synchronous data. To the analysis system 240. The conversion device 230 is independent of the manufacturing system 21 〇 and the host 22 〇. The function of the conversion device 230 is to convert the non-standard data collected by various modes into a data type that can be recognized by the host 220, and transmit the converted data to the host 220. At the same time, in the traditional data synchronization type semiconductor process system, the function of synchronizing resources and events to realize data synchronization must be additionally provided by the host 220. As a result, in the conventional data synchronization type semiconductor process system, the load of the host 220 is increased, and the stability of the system is also deteriorated. Further, in the conventional data synchronization type semiconductor process system, since all the data is collected to the host 220, the network load around the host 220 is increased. The load generated at this time may impair the stability of the system and may hinder the normal operation of the manufacturing process. As a result, the operation of the conventional data synchronous semiconductor process system may be shut down and may even cause serious losses. In addition, traditional data-synchronized semiconductor process systems can cause problems due to the low stability of non-standard data. The low stability of the above-mentioned non-standard data is caused by the conversion of non-standard data into a form recognizable by the host, or by using the network to transfer the converted data, or to temporarily delay the execution of data or event synchronization. The third figure is shown as a schematic diagram of the agent data-sync type semiconductor process system (agent data-Sync type semic〇nduct〇r manufactuHng _ (four). As shown in Figure 3, 'traditional agent f-synchronous semiconductor process system The processing device (10), the host 32, the agent 33G, and the analysis system are included. Because the agent 330 performs data synchronization, the semiconductor processing system can be called, and the proxy data synchronization type system 320 is in the data synchronization type semiconductor process system. The agent 330 is the main material, and the function is provided. The agent 33G collects the standard data and synchronizes the non-standard credit data, and then transmits the data to the analysis system. The end 33^ can be traditional agent _ step type semiconductor system (4) In the middle of the system, the agent lacks a cover/: the system burden of the machine 32〇, but the network burden of the host 320 depends
TW2024PA 二Ϊ代==型資Γ係從製造系統310傳遞至主機咖，然 而3 3 〇，依資料同步觀點，傳統代理資料同步型半導 1303032 ' f) 體製程系統的表現較遜於上述傳 系統。 貝科同步型半導體製程 如同上述’因為標準資料 資料是沒問題的。相反的，因為二二標準’收集與分析標準 而達成資料相容，標準資料二料需要轉換成標準資料 的問題。 〜耗準資料的同步需求會出現較大 特別一提，標準資料是由製 所以很難知道資料所產生的精門H後才透過網路傳送， ^ ^ 時間。^果，报難將標準資料與 非t準貝㈣日讀在相同的臨時輸送軸内。 【發明内容】 之^rru本發明係為解決以上f知的問題，本發明的目的 供-種收集標準資料與非標準資料之线與方法， 與非標準資料可即時同步而非常穩定，且不 ㈣ 體製程系統的操作負載。 f守 本發明的另—目的係提供—種計算機可讀取媒趙，其根據上 述目的以一獨立產品的型式實施。 為達本發a之目的，本發明之—面向係提供—種收集半導體 製程系統之非標準資料之⑽，其藉由連接至—製程設備、一主 機和-分析系統的-資料收集舰器來收集非標準資料，上述系 、、充包括:-執仃程式，儲存在—記憶體中；以及_連接至記憶體 之處理器’其用以執行上述執行程式。其中，根據上述執行程式， 上述處理11指定—狀態變數辨認碼（Status Variable IDentifier; SVID)為一製程設備分析所需之非標準資料、接收一由主機和分 析系統其中之一而來的資料需求訊息、儲存資料需求訊息於記憶 體中、決定資料需求訊息中的一狀態變數辨認碼（SVID)和一收集 TW2024(060622)CRF.doc 9 1303032 事件辨認碼（Collection Event Identifier; CEID)是否已被定義、決 疋一不被製程設備支援的訊息為一非標準資料需求訊息、傳送對 應一標準資料需求訊息的資料需求訊息至製程設備、傳送非標準 >料鸹求汛息至一非標準資料偵測單元，由製程設備接收第一標 準身料、由非標準資料偵測單元收集一非標準資料、轉換非標準 貝料為第二標準資料、同步第二標準資料和第一標準資料，以及 傳送第一標準資料與已同步之第二標準資料資料至主機和分析 系統其中之一。 上述之”標準資料”和，，第一標準資料”包括符合半導體設備 通訊標準（SECS stand㈣之資料。上述之，，非標準㈣，，係指呈有 非特，型式的資料、非由製程設備提供的資料、以及由不同裝置 所獲得的資料，例如感測器或編程邏輯控制器（piOgra_bie nPLC)。此外，上述之”第二標準資料，，包括標準 具有料㈣的資料’例如由非標準資料轉成主機可辨識 收华是，_收集舰^週期性地從非標準資㈣測單; 貝料’並計算非標準資料在第—標準資料到達時心 =州料。在此例中，資料收集魏器 、’=束事件間每間隔週期取樣，以收集 料二 當每間隔週期取樣時，資料收集伺服器每間隔舍換:之 集到之非標準母叫週期還會更新已屯 資料收集伺服器透過非標準資料读測單元， 標準資料。在此例中，非標準資料偵測單元包括集辦 狀態變數的^心取㈣存在製程設備記憶體中 TW2024(060622)CRF.doc 10 1303032 司服器包括傳达/接收模組、—標準資料收集模 組、-非&準資料收集模組以及一資料轉換模組。資料收集舰 器更包括-過濾模組’用以選擇性地傳送部分之第一標準資料或 第二標準資料。以及，資料收集飼服器更包括—通訊切換模組， 用以直接連接在製程設備與主機之間的通訊線路。 根據本發明之另-面向，提供—種收集半導體製程系統之非 標準資料之方法，藉由連接至_製程設借、—主機和—分析李统 的一資料收集舰时收㈣鮮資料，上述方法包括下列步 驟：指定-狀態變數辨認碼（SVID)為―製程設備分析所需之非標 準資料；接收由主機和分析系統其中之一而來的一資料需求訊· 息，館存資料需求訊息於—記憶體中；決定是否已定義資料需求 訊息中的-狀態變數辨認瑪（SVID)和一收集事件辨認碼 (CEID);決定一不被製程設備支援的訊息為一非標準資料需求訊· 息丄傳送對應-標準資料需求訊息的詩需求訊息至製程設備；· 傳达非標準資料需求訊息至一非標準資料偵測單元；由製程設備 接收第-標準資料；由非標準資料谓測單元收集一非標準資料； ，換非標準資料為第三標準f料；以及同步第三標準資料和第一 私準資料’以及傳送第—標準資料與同步之第二標準資料至主機· 和分析系統其中之一。 響 較佳的是’收集非標準資料之方法包括：週期性地由非標準 資料偵測單元收㈣標準諸；計算非標準㈣在—第—標準資 料到達時間之最大值、最小值與平均值；以及處理計算值至與第， 一標準資料同步之非標準資料。 - 收集非標準資料之步驟包括，資料收集飼服器透過在一開始 牛和、、Ό束事件間每間隔週期取樣。換言之，收集非標準資料 之步驟包括，當每間隔週期取樣時，資料收集祠服器會每間隔週 TW2〇24(〇6〇622)CRF.doc 11 [on ( 1303032 期更新已收集到之非標準資料。 傳送非標準資料需求訊息至製程設備之步驟包括，在刪除資 料需求訊息中不被該製程設備支援的訊息之後，傳送非標準資料 需求訊息至製程設備。 上述方法更包括，當由製程設備收到一負認知訊息（ΝΑΚ) 時，取消一被執行的操作。 上述方法更包括，修復關於儲存在記憶體中之資料需求訊 息；以及藉由結合第一標準資料與第二標準資料，以架構一回應 訊息。 上述方法更包括，當資料收集伺服器不正確操作或不穩定 時，在製程設備與主機間直接連接通訊。 根據本發明之再一面向，提供一種計算機可讀取紀錄媒體， 其I儲存一程式，此程式用以執行一收集半導體製程系統之非標 準資料之方法，此方法包括下列步驟··指定一狀態變數辨認碼 (SVID)為-製程設備分析所需之非標準f料；接收由主機和分析 系U之-而來的資料需求訊息；儲存資料需求訊息於記憶體 中，决定疋否已定義貝料需求訊息中的一狀態變數辨認碼（svid) 和-收集事件辨認碼_!));決定—颂製㈣備支援的訊息為 非私準資料需求訊息；傳送對應一標準資料需求訊息的資料需求 Γ息至製程設備；傳送非標準資料需求訊息至一非標準資㈣測 :元；由製程設備接收第-標準資料；由非標準資㈣測單元收 集-非標準資料，·轉換非標準資料為第二標準資料；以及同 = 票準資料和第-標準資料，並傳送第—標準資料與同步之第二 “準資料至主機和分析系統其中之_。 為讓本發明之上述目的、特徵、和優點能更明㈣懂，下文 特舉-較佳實施例，並配合所附圖式，作詳細說明如下： TW2024(060622)CRF.doc 12 1303032 · « 【實施方式】 本發明之實施例配合所附圖式，作詳細說明如下。下文中^ 述到”一物件連接另一物件”包括：兩物件透過中間物件彼此連接 的情況，以及兩物件直接連接的情況。更進一步，與本發明無關 的元件，會在圖式中省略，以使本發明的描述更清楚。在不同的 · 圖式中，相同或類似的元件係沿用相同的標號。 第4圖係繪示本發明一種收集半導體製程系統資料之系統 架構示意圖。 請參照第4圖，本發明收集半導體製程系統資料之系統，係 收集標準資料（typical data)和非標準資料（atypical data)，並在不 鲁 增加系統操作負載的情況下，將這些資料傳送至分析系統。為達 到此目地，上述資料收集系統，包括製程設備41〇、資料收集伺 服器420、主機430、以及分析系統44〇。 · 上述製程設備410係執行一特定的半導體製程。舉例來說，· 此半導體製程包括一清潔製程、一熱製程、一微影製程 (photographing process)、一包裝製程（paekaging process)、一測試 和分析製程。所以，製程設備41〇包括清潔設備、沉積設備、熱 處理設備、曝光設備、顯影設備、包裝設備、以及測試和分析設φ 備。更進一步，當進行半導體製程時，製程設備41〇提供標準資 料至上述育料收集伺服器42〇，其中，標準資料係指符合半導體 ,又備通巩軚準（SECS standard)的資料。此外，製程設備41〇連接 非心準資料偵測單元（atypical如仏d咖⑷⑽unit)(未顯示），上 j非軚準貝料偵測單元係提供非標準資料至上述資料收集伺服 · 裔420，其中，非標準資料係指不符合半導體設備通訊標準的資 料。 在上例中，非標準資料偵測單元包括一裝置，用以測量製程TW2024PA Ϊ2 == Γ 传递 传递 = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = = system. The Beca Synchronous Semiconductor process is as described above because the standard data is no problem. On the contrary, because of the data compatibility between the two standards' collection and analysis standards, the standard data needs to be converted into standard data. ~ The need for synchronization of data consumption will be a big mention. The standard data is based on the system. It is difficult to know the data generated by the door H and then transmitted through the network, ^ ^ time. ^, it is difficult to report the standard data and the non-t quasi-before (four) day in the same temporary transport axis. SUMMARY OF THE INVENTION The present invention is directed to solving the above problems, and the object of the present invention is to provide a line and method for collecting standard data and non-standard data, which can be synchronized with non-standard data and is very stable, and is not (iv) Operational load of the institutional process system. Another object of the present invention is to provide a computer readable medium, which is implemented in the form of a separate product in accordance with the above objects. For the purposes of the present invention, the present invention provides a non-standard material for collecting semiconductor process systems (10) by means of a data collection vessel connected to a process device, a host and an analysis system. Collecting non-standard data, the above-mentioned system, including: - the execution program, stored in the - memory; and - the processor connected to the memory - is used to execute the above execution program. According to the above execution program, the processing 11 specifies that the Status Variable IDentifier (SVID) is a non-standard data required for analysis of a process device, and receives a data request from one of the host and the analysis system. The message, the stored data request message in the memory, a state variable identification code (SVID) in the data request message, and a collection TW2024 (060622) CRF.doc 9 1303032 Event Identification Identifier (CEID) have been Defining, determining that a message not supported by the process device is a non-standard data request message, transmitting a data request message corresponding to a standard data request message to a process device, transmitting a non-standard > requesting a message to a non-standard data The detecting unit receives the first standard body by the processing device, collects a non-standard data by the non-standard data detecting unit, converts the non-standard bedding material into the second standard data, synchronizes the second standard data and the first standard data, and Transfer the first standard data and the synchronized second standard data to one of the host and the analysis system. The above-mentioned "standard data" and, "first standard data" include information conforming to the semiconductor device communication standard (SECS stand (4). The above, non-standard (4), refers to the non-special, type of information, not provided by the process equipment Information, as well as information obtained by different devices, such as sensors or programmable logic controllers (piOgra_bie nPLC). In addition, the above-mentioned "second standard data, including the standard material (4) data", for example, from non-standard data Turn into a host to identify the collection, _ collection ship ^ periodically from non-standard capital (four) test; bedding 'and calculate non-standard data in the first - standard data arrival heart = state material. In this case, the data Collecting the Wei instrument, sampling every interval between the '= beam events, to collect the material. When sampling every interval, the data collection server exchanges every interval: the non-standard mother call period is also updated to update the data collection. The server passes the non-standard data reading unit, standard data. In this example, the non-standard data detection unit includes the collection state variable (4) the presence of the device memory. TW2024(060622)CRF.doc 10 1303032 The server includes a communication/reception module, a standard data collection module, a non-standard data collection module, and a data conversion module. The data collection vehicle further includes - The filter module is configured to selectively transmit a portion of the first standard data or the second standard data. The data collection and feeding device further includes a communication switching module for directly connecting the communication between the process device and the host computer. In accordance with another aspect of the present invention, a method of collecting non-standard data of a semiconductor process system is provided by means of a data collection ship connected to a process, a host, and an analysis of Li Tong. The above method comprises the steps of: specifying a state variable identification code (SVID) as non-standard data required for process device analysis; receiving a data request message and information from one of the host and the analysis system, and storing the library data The demand message is in the memory; it is determined whether the -state variable identification number (SVID) and a collection event identification code (CEID) in the data requirement message have been defined; The message of assistance is a non-standard data request message, information transfer message corresponding to the standard data request message to the process equipment; • communication of non-standard data request message to a non-standard data detection unit; receiving by the process equipment - standard data; non-standard data pre-test unit collects a non-standard data; non-standard data for the third standard f material; and synchronous third standard data and first private data 'and the first-standard data and synchronization The second standard data is one of the host and analysis systems. It is better that the method of collecting non-standard data includes: periodic collection by non-standard data detection units (four) standards; calculation of non-standard (four) in - The maximum, minimum, and average values of the first-standard data arrival time; and the non-standard data that is processed to the calculated value to be synchronized with the first, standard data. - The step of collecting non-standard data includes the data collection and feeding device sampling at intervals between the beginning of the cattle and the event. In other words, the step of collecting non-standard data includes, when sampling every interval, the data collection server will be TW2〇24 (〇6〇622) every week. CRF.doc 11 [on (1303032 update has been collected) Standard data. The step of transmitting a non-standard data request message to the process device includes transmitting a non-standard data request message to the process device after deleting the message that is not supported by the process device in the data request message. The above method further includes when the process is When the device receives a negative cognitive message (ΝΑΚ), cancels an executed operation. The above method further includes repairing a data demand message stored in the memory; and by combining the first standard data with the second standard data, The above method further includes directly connecting communication between the process device and the host when the data collection server is incorrectly operated or unstable. According to still another aspect of the present invention, a computer readable recording medium is provided. I store a program for performing a method of collecting non-standard data of a semiconductor process system The method includes the following steps: specifying a state variable identification code (SVID) for the non-standard f material required for the analysis of the process device; receiving the data demand message from the host and the analysis system U; storing the data demand message In the memory, it is determined whether a state variable identification code (svid) and a collection event identification code _!) in the bedding material demand message are defined; the decision-type (four) backup support message is a non-private data request message. Transmitting the data requirements corresponding to a standard data request message to the process equipment; transmitting the non-standard data demand message to a non-standard resource (4) measurement: element; receiving the first-standard data by the process equipment; collecting by the non-standard resource (four) measurement unit - Non-standard data, · Convert non-standard data to second standard data; and the same = ticketing data and first-standard data, and transmit the first - standard data and synchronization of the second "quasi-data to the host and analysis system" The above described objects, features, and advantages of the present invention will become more apparent from the following description of the preferred embodiments of the invention. 1303032 « «Embodiment of the Invention The embodiments of the present invention are described in detail below with reference to the accompanying drawings. In the following, "an object connected to another object" includes: a case where two objects are connected to each other through an intermediate object, and two In the case where the articles are directly connected, the components that are not related to the present invention will be omitted in the drawings to make the description of the present invention clearer. In the different drawings, the same or similar components follow the same reference numerals. Figure 4 is a schematic diagram of a system architecture for collecting semiconductor process system data according to the present invention. Referring to Figure 4, the system for collecting semiconductor process system data of the present invention collects common data and non-standard data (atypical) Data), and transfer the data to the analysis system without increasing the operating load of the system. To achieve this, the above data collection system includes a process device 41, a data collection server 420, a host 430, and an analysis system 44A. The process device 410 described above performs a particular semiconductor process. For example, the semiconductor process includes a cleaning process, a thermal process, a photographing process, a paekaging process, a test and analysis process. Therefore, the process equipment 41 includes cleaning equipment, deposition equipment, heat treatment equipment, exposure equipment, development equipment, packaging equipment, and test and analysis equipment. Further, when the semiconductor process is performed, the process equipment 41 provides standard data to the above-mentioned feed collection server 42A, wherein the standard data refers to the information conforming to the semiconductor and the SECS standard. In addition, the process device 41 is connected to a non-precise data detecting unit (atypical, such as 仏d coffee (4) (10) unit) (not shown), and the non-standard material detecting unit provides non-standard data to the above data collecting servo 420 Among them, non-standard data refers to materials that do not meet the communication standards of semiconductor devices. In the above example, the non-standard data detection unit includes a device for measuring the process
TW2024PA 13 1303032 :備4U)之非標準資料，或取得儲存在製 4 例來說，上述非標準資料偵測單元包括測量二^ 二I：力、時間等的感應H。如上所述，非標準資㈣測單元 基本上用以偵測非標準資料，此非標準資料並非直接由製程設備 提供’而是由例如直接接觸-感測器或編程邏輯㈣ 來。 製程設備410產生一反應以回應由主機43〇而來的資料請 求。更進-乡，製程設冑41(M則量符合半導體設備通訊標準的標 準資料，並提供所測量的標準資料至資料收集伺服器42〇。為達 到此目地，製程設備410可以包含一編程邏輯控制器 (programmable logic contr〇iier; PLC)(未顯示）、一測量裝置（未顯 示）以及一通訊裝置（未顯示）。 上述編程邏輯控制器係一可自動控制和監控製程設備41〇 之元件。編程邏輯控制器可控制測量裝置和通訊裝置，以及提供 已測量之資料給資料收集伺服器420，以回應由資料收集伺服器 420傳來的訊息。在此例中，測量裝置例如一裝設在製程設備41〇 上之元件，其測量標準資料例如溫度或壓力。通訊裝置例如一配 置，其將製程設備410與資料收集伺服器42〇透過一 Rs_232訊 號線或本地區域網路（Local Area Network; LAN)來連接。 資料收集伺服器420透過TCP/IP通訊方法以接收由主機 430及/或分析系統440來的資料請求。在此例中，資料收集伺服 器420由主機430接收到的訊息，包括用於製造過程直接需要的 請求訊息。相對的，資料收集伺服器420由分析系統44〇接收到 的说息’包括用於分析以改進製造效率直接需要的請求訊息。 當與製程設備410通訊時，資料收集伺服器420收集標準資 料’而當與非標準資料债測單元通訊時，資料收集飼服器420收 TW2024PA 14 1303032 集非払準貝料。更進一步，資料收集伺服器42〇傳送所需的資料 :主機430與刀析系統44〇，以回應主機㈣與分析系统柳的 明求在此例中’當利用過濾功能增加或除去訊息後，資料收集 伺服器420可以傳迸已收集之標準和非標準資料。 一、主機430可控制製程設備41〇，使得製程設備4i〇執行一特 疋半導體製程。此時，根據半導體設備通訊標準，主機43〇與製 私„又備410父換矾息。在此實施例中，主機43〇不僅收集非標準 資料，貫際上，主機430有包括同步標準資料與非標準資料之功 能。 主機430連接到儲存大量製程設備41〇所需之製程資料的資 料庫。在此例中，主機430可以是主機元件或主機電腦，連接到 製程設備上使之能操作至少部分的製程設備。 分析系統440可接收製程設備410的狀態，以及改變製程設 備410從資料收集伺服器42〇測量到的資料。 與ΕΑΡ資料同步系統以及與傳統代理資料同步系統比較起 來’上述本發明之資料收集系統可以被視為一，，内部資料同步系 系先，，(inter data-sync type system) 〇 同時，上述本發明資料收集系統可進一步包括使用者介面以 及操作者介面。在此例中，使用者介面包括一電腦，藉由此電腦， 操作者或分析者分析製程資料與系統問題所在的緣由。操作者介 面包括一電腦，藉由此電腦，操作者可檢查儲存在主機資料庫的 製程資料，以及執行一半導體製程。 根據本發明實施例所描述之架構，當製程設備被監控時，必 要的標準與非標準資料會被收集起來，而半導體製程系統的網路 負擔並不會增加太大。所以，半導體系統可以穩定地維持和運作。 如上所述，本發明提供一系統，其不僅可用以收集標準資 TW2024PA 15 1303032 0.>” 他ι 料，亦可同時收集與標準資料同步之非標準資料。換句話說，當 分析系統被引進用來有效操作半導體製程系統時，本發明之資料 收集系統能夠有效地收集標準與非標準資料，此系統可在不增加 半導體製程系統操作負擔的情況下，即時彼此同步以及維持穩定 性。 u 貝料收集伺服器的同步方法詳述如下。第5a圖和第5b圖係 繪示根據第4圖資料收集系統的資料同步方法之示意圖。 =第5a圖所示，本發明實施例之資料收集系統，可藉由在 次特疋期間内每固^間隔取樣’以收集非標準資料。特別地，此 資料收集伺服器，可藉由對非標準資料偵測單元提出詢問 (qUerieS)Ql至Q5,在一特定期間内從開始事件（start events T1 至結f事件點T2，以取樣非標準資料，其中，上述非標準資料 ^則單7〇連接至製程設備。更進—步，當從製程設備接收到標準 資料時，資料收㈣服！！可計算已取樣之非標料料的最大值、 最小值與平均值，將上述之非標準資料值再與標準資料作同步處 理，以及將標準資料與非標準資料一起傳送至分析系統。 -如第5b圖所示，本發明實施例之f料收㈣統，當每固定 間隔取樣非標準資料時，可藉由更新在固定間隔取樣的資料，以 收集非標準資料。特別地，當取樣非標準資料時，此資料收集伺 服器’可藉由對非標準資料偵測單元提出詢問Q1 i Q7,每固定 間隔點T卜T2# T3更新取樣的資料，以取樣 疋 >上述非標準資㈣測單元連接至製程設備。更進—/，ζ從、中’ 接收到非標準資料時，t料收㈣服时計算已取樣之 準貝料的最大值、最小值與平均值，將上述之非標準資料值與^ 收到的標準資料作同步處理，以及將標準資料與 二 傳送至分析系統。 貝升一起 TW2024(060622)CRF.doc 16 1303032 藉由上述之結構，本發明實施例之資料收集系統可及時收集 非標準資料，且將標準資料與非標準資料作同步處理，以及傳送 標準資料與非標準資料至分析系統。所以，根據本發明，可不用 增加半導體製程系統的操作負擔，就有效收集到可靠的資料。此 外’亦可在需要偵測溫度過熱與壓力過大之情況時，有效地應用 本發明之資料收集系統。 第6圖係繪示根據第4圖資料收集系統的資料收集伺服器之 示意圖。 請參照第6圖，資料收集伺服器420包括記憶體系統421、 至少一中央處理單元（CPU)422、輸入系統427、以及通訊系統 429,其中，中央處理單元422連接至記憶體系統421，且適合執 行南速操作。 中央處理早元422包括算術邏輯單元424、登錄器426、以 及控制器428，其中，算術邏輯單元424用以執行計算功能，登 錄器426用以暫時儲存資料和指令，而控制器428用以控制資料 收集伺服器420之操作。中央處理單元422是任一種處理器，例 如數位Alpha晶片，或是MIPS科技、NEC、IDT和西門子（Siemens) 之MIPS處理器、或是英特爾（Intel)、Cyrix、超微（AMD)和Nexgen 之x86處理器、以及IBM和摩托羅拉（Motorola)之PowerPC等。 記憶體系統421 —般包括高速主記憶體423、輔助記憶體 425、以及利用電、磁、光或其它形式來儲存資料的元件。其中， 上述高速主記憶體423具有儲存媒體形式，例如隋機存取記憶體 (RAM)或唯讀記憶體（ROM)等，上述辅助記憶體425具有長期保 存之儲存媒體形式’例如軟碟、硬碟、磁碟（magnetic disk)、光 碟（compact disk)以及快閃記憶體等。更進一步，高速主記憶體 423可以包括影像記憶體，其可透過影像元件來顯示影像。 TW2024PA 17 1303032 輸入元件427包括鍵盤和滑鼠，滑鼠包括物理變換器 (physical transducer)，例如觸控螢幕或麥克風。通訊元件429包 括通訊介面和無線傳送/接收介面，上述無線傳送/接收介面係用 以與製程設備、主機和分析系統作通訊。 第7圖係繪示根據第6圖資料收集伺服器的模組結構示意 圖。 請參照第7圖，資料收集伺服器包括傳送/接收處理模組 741、標準資料收集模組743、非標準資料收集模組745、資料轉 換模組（data conversion module)747、以及過渡處理模組（filtering processing module)749。其中，傳送/接收處理模組741係用以處 理與製程設備、主機和分析系統之通訊，標準資料收集模組743 係由製程設備收集標準資料，非標準資料收集模組745係由與製 程設備連接之非標準資料偵測單元以收集非標準資料，資料轉換 模組747係用以將已收集到的非標準資料轉換為一能由主機、分 析系統或標準資料所辨識之特定型式的資料，過濾處理模組749 係用以處理由製程設備與非標準資料偵測單元收集到的資料、以 及選擇性地傳送已處理之資料至主機和/或分析系統。本發明資料 收集伺服器之整個架構係對應各應用模組詳述如下。 資料收集伺服器可使用不同操作系統（OS)中之任一種，以作 為本系統之操作系統。例如一操作系統，其可藉由提供高等命令 (high-level instruction)至應用程式介面（Application Program Interface;API)，來控制應用模組740之操作。 資料收集伺服器包括高等命令處理單元710,其根據應用程 式介面（API)700提供之高等命令，以辨識各個應用模組，以及可 解碼高等命令與提供已解碼之高等命令至對硬的應用模組。資料 收集伺服器還包括應用模組控制單元720，其根據高等命令處理 TW2024PA 18 1303032 單元710所提供之高等命令，以控制應用模組740之操作。換言 之，高等命令處理單元710根據應用程式介面（API)700提供之高 等命令，以決定對應之應用模組740是否存在，若對應之應用模 組740存在，則解碼高等命令成一可被應用模組740辨識之命 令，以及傳送上述命令至一對應之映像單元（mapping unit)或控制 訊息的傳遞。應用模組控制單元720包括映像單元721、723、 725、727 寺口 729，與介面 722、724、726、728 #口 730，以分別用 於應用模組 741、743、745、747 和 749 ° 傳送/接收處理模組映像單元721接收由主機和分析系統而 來之資料要求訊息，透過網路傳送標準資料和/或非標準資料至主 機或分析系統，以及與製程設備和非標準資料偵測單元溝通以收 集標準資料和非標準資料。傳送/接收處理模組映像單元721接收 高等命令處理單元710之高等命令，映像（map)高等命令成一能 由傳送/接收處理模組741處理之元件命令（device-level instruction)，以及透過傳送/接收處理模組介面722，提供上述元 件命令至傳送/接收處理模組741。若是需要的話，傳送/接收處 理模組741可包括一物理轉換介面，以直接溝通連接製程設備與 主機。在此例中，物理轉換介面包括一通訊切換介面 (communication switching interface) » 例如一開關（switch) 〇 標準資料收集模組映像單元723接收由主機和/或分析系統 而來之資料要求訊息，在記憶體系統儲存上述資料要求訊息，解 讀上述資料要求訊息，以及收集從製程設備來之標準資料。為達 此目地，標準資料收集模組映像單元723接收高等命令處理單元 710之高等命令以收集非標準資料，映像高等命令成一能由標準 資料映像模組723辨識之元件命令，以及透過標準資料收集模組 介面724，提供上述元件命令至標準資料收集模組743。 TW2024PA 19 1303032 非標準資料收集模組映像單元725決定是否訊息（其並不限 定用以分析製程設備）被包含於由主機和/或分析系統傳來之資料 要求訊息中。此外，非標準資料收集模組映像單元725還從與製 程設備連接之非標準資料偵測單元，來收集非標準資料。為達此 目地，非標準資料收集模組映像單元725接收高等命令處理單元 710之高等命令以收集非標準資料，映像高等命令成一能由非標 準資料收集模組745辨識之元件命令，以及透過非標準資料收集 模組介面726，提供上述元件命令至非標準資料收集模組745。 資料轉換模組映像單元727能將由非標準資料偵測單元收 集到的非標準資料，轉換成標準資料或能由主機或分析系統辯識 之資料。為達此目地，資料轉換模組映像單元727接收一高等命 令，其能轉換非標準資料成標準資料或從高等命令處理單元710 來的資料，映像高等命令成一能由資料轉換模組747辨識之元件 命令，以及透過資料轉換模組介面728，提供上述元件命令至資 料轉換模組747。 當從製程設備和非標準資料偵測單元收集的資料傳送至主 機或分析系統時，過濾處理模組映像單元729選擇性的處理已收 集的資料，只允許傳送不需要的和被要求的資料。為達此目地， 過濾處理模組映像單元729接收一高等命令，其能選擇性的將高 等命令處理單元710已收集或產生的資料，傳送至主機和/或分析 系統。映像高等命令成一能由過濾處理模組749辨識之元件命 令，以及透過過濾處理模組介面730，提供上述元件命令至過濾 處理模組749。舉例而言，當主機或分析系統從製程設備和非標 準資料彳貞測單元要求區塊資料（block-based data)時，上述的架構 可從資料中去除不需要的資料，或是選擇性的傳送已收集或產生 的資料至主機和/或分析系統，因此可避免網路負擔的增加，以及 TW2024PA 20 1303032 避免製造線穩定性的降低。 如上所述’本發明實施例之資料收集系統係以下列方式架構 起來，在主機和製程設備之間架設獨立的資料收集伺服器，以在 主機和製程設備之間傳達訊息，此外，還執行標準資料的收集、 非標準資料的收集、資料的轉換以及資料的同步。因此，根據本 發明，標準資料和非標準資料能及時被收集，不會有大量的時間 延遲，使得資料同步的問題可被解決。更進一步，既然主機要求 不改變，系統負擔的問題一此問題經常在傳統的半導體製程系統 中發生一可以被克服。除此以外，本發明還可透過以收集資料與 内部資料同步結構（inter data-sync type configuration)之過濾，以 控制避免不需要的資料傳送至主機或分析系統，使得網路負擔增 加的問題——此問題經常在傳統的半導體製程系統中發生——可 以被解決。 請參照第8圖，收集系統的訊號流（signal flow)描述如下。 多數半導體製程設備使用GEM(Generic model for communications and control of Manufacturing equipment)通訊協 定，來描述主機、分析系統與製程設備之間的訊息種類與意義。 因此，支援GEM通訊協定的分析系統根據GEM通訊協定收集資 料。當製程設備不支援GEM通訊協定或存在非標準型資料，增 加支援GEM通訊協定的流程是需要的。此發明實施例提供用於 上述情形的處理流程表。 在本發明實施例的資料收集系統中，主機和/或分析系統根 據GEM協定透過資料收集伺服器收集藉由狀態變數辨認碼 (Status Variable IDentifier; SVID)來定義所需要的資訊。大多數狀 態變數辨認碼（SVID)係藉由報告辨認碼（Report IDentifier; RPTID)來收集管理。更進一步，主機和/或分析系統定義與管理 TW2024PA 21 1303032 收集事件辨認碼（Collection Event IDentifier; CEID) ’ 製造系統 < 利用此碼通知事件的發生’例如緊急事件。在此例中’事件辨認 石馬通常參照非主機要求的辨識碼來定義一事件的發生。基於此目 的，資料收集系統定義此訊息規則，使訊息可以透過分析系統滅1 此交換上述已執行的功能。GEM協定定義了一些標準訊息’參 照第8圖，這些訊息與資料收集伺服器中的資料同步相關° 第8圖係顯示根據此發明實施例之資料收集系統的訊號流° 參照第8圖中S1F3與S1F4訊息，分別代表著主機或分析 系統用於查詢關於一從製程設備傳來的被選擇狀態變數辨認碼 (ISVID)，與一回應前者訊息的製造訊息。舉例來說，製程設備第 一操作室（chamber) 1的溫度被設定成SVID 1與第二操作室2的 壓力被設定為SVID 2,當必須知道第一操作室1的溫度與第二操 作室2的壓力時，主機在訊息S1F3中編寫SVID 1與SVID 4 ’ 然後傳送訊息S1F3到製程設備，製程設備將第一操作室1的溫 度與第二操作室2的壓力加在訊息S1F4中，然後傳送訊息S1F4 回到主機或分析系統。 訊息S2F23與S2F24分別代表發出追蹤命令給製程設備且 使其周期性回報不同狀態變數辨認碼（SVID)的訊息，以及製程設 備用以回應前述訊息的訊息。當使用S1F3與S1F4訊息來收集狀 態變數辨認碼有困難時，就可使用訊息S2F23與S2F24。例如： 主機藉由訊息S2F23命令製程設備以每秒一千次的週期來偵測 狀態變數辨認碼SVID 1，SVID 2，SVID 3，SVID 4，並且製程設備 利用訊息S2F24來回應訊息S2F23。 訊息S2F33與S2F34分別代表使用狀態變數辨認碼包（a bundle of SVIDs)以重新定義與刪除報告辨認碼（RPTID)的訊息， 以及製程設備用以回應前述訊息的訊息。 TW2024PA 22 1303032 訊息82?35與82?36分別代表連接報告辨認碼包（&131111(14 of RPTIDs)到收集事件辨認碼（CEID)或刪除一已存在連線的訊 息，以及製程設備用以回應前述訊息的訊息。 訊息S2F37與S2F38分別代表開啟（enable)或關閉（disable) 收集事件辨認碼（CEID)的訊息，以及製程設備用以回應前述訊息 的訊息。 換言之，當製程設備報告一事件時，主機可定義與收集事件 辨認碼（CEID)有關的報告辨認碼（RPTID)以及與報告辨認碼 (RPTID)有關的狀態變數辨認碼（SVID)。在此例中，收集事件辨 認碼（CEID)以狀態變數辨認碼包來傳送，這些碼在事件發生時就 已存在且互相結合。此種變數辨認碼包被被稱為一報告辨認碼 (RPTID)。 訊息S2F33係指主機用來定義多數狀態變數辨認碼為一報 告辨認碼以及刪除報告辨認碼的訊息，而訊息S2F34係指製造系 統回應前述訊息的訊息。訊息S2F35係指連接多數狀態變數辨認 碼到收集事件辨認碼或不連接狀態變數辨認碼到收集事件辨認 碼的訊息，訊息S2F36係指製造系統回應前述訊息的訊息。訊息 S2F37係指依據主機是否對一特定收集事件辨認碼作報告而啟動 或關閉一特定事件的訊息，而訊息S2F38係指製造系統回應前述 訊息的訊息。 訊息S6F2與S6F1分別代表引發狀態變數(SV)或狀態變數 辨認碼（SVID)的週期報告，以回應追蹤指令的訊息，以及製造系 統用以回應前述訊息而產生的報告訊息。例如，訊息S6F1是製 造系統用於回應主機的追蹤指令，以增加狀態變數辨認碼數值的 訊息。訊息S6F2係主機用以回應製程設備訊息S6F1的訊息。 訊息S6F12與S6F11分別代表引發製造系統報告一具有特 TW2024PA 23 1303032 定收集事件辨認碼之事件的訊息，以及製造系統用於回應前述訊 息而產生的報告訊息。當製程設備需要對主機報告事件時，就會 使用到成息S6F12與S6F11。例如，在收集事件辨認碼CEID i 被定義為程序起頭，收集事件辨認碼cmD2被定義為卡匿載入 完成的例子中，製程設備傳送訊息S6FU到一寫入收集事件辨認 碼的主機，而主機傳送S6F12到製程設備以回應訊息S6fu。 訊息S6F15與S6F16分別代表主機得知一關於收集事件辨 認，與的報告辨認碼，和得知從製造系統查詢收集事件辨認碼的 狀悲變數的訊息，以及製造系統用於回應前述訊息而產生的報告 訊息。換句話講，訊息S6F15係指當主機要得知關於一特定收集 事件辨認碼的報告辨認碼與狀態變數辨認碼數值時，主機用於編 寫一收集事件辨認碼，以及傳遞注意事項的訊息。訊息8矸16為 製造系統用於編寫關於-特定收集事件辨認碼的報告辨認碼與 狀態變數辨認碼數值，與傳遞一回應前述訊息的注意事項的訊 息。 、 訊息S 6 F! 9與s 6 F 2 0分別代表主機用於得知從製造系統查 詢的報告辨認碼中所含狀態變數辨認碼的訊息，與製造系統回應 前述訊息而產生報告的訊息。此例中狀態變數代表狀態變數㈣ 碼的數值。 同時，上述訊息中的，’S”與”F”分別代表，，串流，，(stream)與”功 能’’(function)，用以將主機與製造系統間交換訊息進行分類。例 如，訊息S6F19代表對應到串流6與功能19的訊息。更進一步， 上述訊息係由主機透過資料收集伺服器傳送到製程設備的資料 需求訊息，而所收集的資料則從製程設備透過資料收集伺服器傳 到主機且/或分析系統。 / 如上所述，根據本發明的非標準資料收集系統，由資料收集 TW2024PA 24 1303032 伺服益所收集到的資料可被同步化與處理過，使得這些資料如 製程設備所提供的一樣。 以下將用於上述非標準資料收集系統的非標準資料收集方 法彳田述於下。第9圖至第12圖係繪示根據本發明實施例之半 ‘體製程系統收集非標準資料之方法流程圖。 參照第9圖，首先，資料收集伺服器指定一由分析系統所定 義的狀您變數辨認碼（SVID)為一製程設備在階段si〇分析時所* 要=到的非標準資料，其中，上述狀態變數辨認碼係由分析系: 所疋義。之後’在階段S12，資料收集伺服器接收到從主機且/ 或分析系統發出的一資料需求訊息。在階段S14，資料收集伺服 器將從主機且/或分析系統發出的資料需求訊息儲存在記憶體 中。接著，在階段S16，資料收集伺服器決定在製程設備所定義 的資料需求訊息中’是否包含狀態變數辨別碼與收集事件辨認碼 (CEID)。製程設備所不支援㈣訊，係被解釋為階段川的非^ 準資料要求。之後在階段S2G，進行非標準f料的㈣。同時: 階段S22，—從資料需求訊息中，資料收集伺服器移除製程設備所 不支援的資訊。更進—步，在階段S24，關於非標準資料的要求 訊息被傳送到製程設備。 參照第10圖’在階段S26，資料收集伺服器經由連接到製 程設備的非鮮資料_單元來收㈣標準料。在此例中使 用第5a圖與第5b圖所描述的方法以完成非標準資料的收集。 之後，在階段S28,資料收㈣服器收集來 準資料。此階段S30,當禪準資㈣㈣ ^…田铩旱貝枓被收集時’資料收集伺服器會 處理非Μ讀，使非標準賴與標準㈣進㈣^例如者 回應訊息回應至少訊息S6F1S6F11S6F15 s6fi9與則其中 任-訊息時，而這些訊息包含當執行非標準資料收集時從製程設TW2024PA 13 1303032: Prepare 4U) non-standard data, or obtain storage in 4 cases, the above non-standard data detection unit includes measuring H2: force, time, etc. As mentioned above, the non-standard (four) measurement unit is basically used to detect non-standard data, which is not directly provided by the process equipment' but by, for example, direct contact-sensor or programming logic (4). The process device 410 generates a response in response to a request for data from the host computer 43. Further into the township, the process setting 41 (M is in accordance with the standard data of the semiconductor device communication standard, and provides the measured standard data to the data collection server 42. To achieve this purpose, the process device 410 can include a programming logic A controller (programmable logic contr〇iier; PLC) (not shown), a measuring device (not shown), and a communication device (not shown). The above programming logic controller is a component that can automatically control and monitor the process device 41. The programming logic controller can control the measuring device and the communication device, and provide the measured data to the data collecting server 420 in response to the message transmitted by the data collecting server 420. In this example, the measuring device is installed, for example. An element on the process device 41 that measures standard data such as temperature or pressure. The communication device, for example, a configuration that passes the process device 410 and the data collection server 42 through an Rs_232 signal line or local area network (Local Area Network) ; LAN) to connect. The data collection server 420 is received by the host 430 and/or the analysis system 440 via a TCP/IP communication method. In this example, the data collection server 420 receives the message received by the host 430, including the request message that is directly needed for the manufacturing process. In contrast, the data collection server 420 is received by the analysis system 44. The information 'includes request information for analysis directly to improve manufacturing efficiency. When communicating with the process device 410, the data collection server 420 collects standard data' and when communicating with the non-standard data debt measurement unit, the data collection feeder 420 receives TW2024PA 14 1303032 set non-standard materials. Further, the data collection server 42 transmits the required data: the host 430 and the knife analysis system 44〇, in response to the host (4) and the analysis system Liu's request in this case When the message is added or removed by the filtering function, the data collection server 420 can transmit the collected standard and non-standard data. 1. The host 430 can control the process device 41〇, so that the process device 4i can execute a special semiconductor. At this time, according to the communication standard of the semiconductor device, the host 43 and the private device are replaced by the suffocation. In this embodiment, the host 43〇 Collecting only non-standard data, the host 430 has a function of including synchronous standard data and non-standard data. The host 430 is connected to a database of process materials required to store a large number of process devices 41. In this example, the host 430 The host component or host computer can be coupled to the process device to enable operation of at least a portion of the process device. The analysis system 440 can receive the status of the process device 410 and change the data measured by the process device 410 from the data collection server 42. Compared with the data synchronization system and the traditional proxy data synchronization system, the above-mentioned data collection system of the present invention can be regarded as one, the internal data synchronization system first, (inter data-sync type system) The data collection system of the present invention can further include a user interface and an operator interface. In this example, the user interface includes a computer by which the computer, operator or analyst analyzes the cause of the process data and system problems. The operator interface includes a computer by which the operator can review process data stored in the host database and perform a semiconductor process. According to the architecture described in the embodiment of the present invention, when the process equipment is monitored, the necessary standard and non-standard data are collected, and the network burden of the semiconductor process system does not increase too much. Therefore, the semiconductor system can be stably maintained and operated. As described above, the present invention provides a system which can be used not only to collect standard resources TW2024PA 15 1303032 0.>", but also to collect non-standard data synchronized with standard data. In other words, when the analysis system is When introduced to effectively operate a semiconductor process system, the data collection system of the present invention can efficiently collect both standard and non-standard data, and the system can instantly synchronize with each other and maintain stability without increasing the operational burden of the semiconductor process system. The synchronization method of the material collection server is described in detail below. Figures 5a and 5b are diagrams showing the data synchronization method of the data collection system according to Fig. 4. = Figure 5a, data collection of the embodiment of the present invention The system can collect non-standard data by sampling every solid interval period. In particular, the data collection server can make an inquiry (qUerieS) Ql to Q5 by using a non-standard data detection unit. , from the start event (start events T1 to knot f event point T2, to sample non-standard data, wherein the above non-standard data ^ is 7 〇 Connect to the process equipment. Further, when receiving standard data from the process equipment, the data is received (4)!! The maximum, minimum and average values of the sampled non-standard materials can be calculated. The standard data values are then processed synchronously with the standard data, and the standard data is transmitted to the analysis system along with the non-standard data. - As shown in Figure 5b, the f-receiving (four) system of the embodiment of the present invention, when sampling at a fixed interval Standard data can be collected by updating the data sampled at regular intervals to collect non-standard data. In particular, when sampling non-standard data, the data collection server can ask Q1 for non-standard data detection unit. i Q7, every fixed interval T T T2# T3 update the sampled data to sample 疋> The above non-standard (four) measurement unit is connected to the process equipment. Further -/, ζ from, and 'when receiving non-standard data , t receiving (four) service time to calculate the maximum, minimum and average value of the sampled quasi-bean material, the above non-standard data values and ^ received standard data for simultaneous processing, and standard data and second pass It is sent to the analysis system. BB2024 (060622) CRF.doc 16 1303032 With the above structure, the data collection system of the embodiment of the present invention can collect non-standard data in time, and synchronize the standard data with the non-standard data. And transfer standard data and non-standard data to the analysis system. Therefore, according to the present invention, reliable data can be efficiently collected without increasing the operational burden of the semiconductor process system. In addition, it is also necessary to detect temperature overheating and excessive pressure. In the case of the present invention, the data collection system of the present invention is effectively applied. Fig. 6 is a schematic diagram showing the data collection server of the data collection system according to Fig. 4. Referring to Fig. 6, the data collection server 420 includes a memory system 421. At least a central processing unit (CPU) 422, an input system 427, and a communication system 429, wherein the central processing unit 422 is coupled to the memory system 421 and is adapted to perform a south speed operation. The central processing early element 422 includes an arithmetic logic unit 424 for performing computing functions, a register 426 for temporarily storing data and instructions, and a controller 428 for controlling the data and instructions. The operation of the data collection server 420. The central processing unit 422 is any type of processor, such as a digital Alpha chip, or a MIPS processor from MIPS Technologies, NEC, IDT, and Siemens, or Intel, Cyrix, AMD, and Nexgen. X86 processors, as well as IBM and Motorola PowerPC. The memory system 421 typically includes a high speed main memory 423, an auxiliary memory 425, and components that store data using electrical, magnetic, optical, or other forms. The high-speed main memory 423 has a storage medium format, such as a memory access memory (RAM) or a read-only memory (ROM). The auxiliary memory 425 has a long-term storage medium format such as a floppy disk. Hard disk, magnetic disk, compact disk, and flash memory. Further, the high speed main memory 423 can include an image memory that can display an image through the image element. TW2024PA 17 1303032 Input component 427 includes a keyboard and a mouse. The mouse includes a physical transducer such as a touch screen or a microphone. The communication component 429 includes a communication interface and a wireless transmit/receive interface for communicating with the processing device, the host, and the analysis system. Fig. 7 is a schematic view showing the structure of a module for collecting a server according to Fig. 6. Referring to FIG. 7, the data collection server includes a transmission/reception processing module 741, a standard data collection module 743, a non-standard data collection module 745, a data conversion module 747, and a transition processing module. (filtering processing module) 749. The transmission/reception processing module 741 is configured to process communication with the process equipment, the host computer, and the analysis system. The standard data collection module 743 collects standard data by the process equipment, and the non-standard data collection module 745 is composed of the processing equipment. A non-standard data detection unit is connected to collect non-standard data, and the data conversion module 747 is used to convert the collected non-standard data into a specific type of data that can be recognized by the host, the analysis system or the standard data. The filtering processing module 749 is configured to process the data collected by the process device and the non-standard data detecting unit, and selectively transfer the processed data to the host and/or the analysis system. The entire architecture of the collection server of the present invention corresponds to each application module as detailed below. The data collection server can use any of a variety of operating systems (OS) as the operating system of the system. For example, an operating system can control the operation of the application module 740 by providing a high-level instruction to an Application Program Interface (API). The data collection server includes a high-level command processing unit 710 that recognizes each application module according to a high-level command provided by an application interface (API) 700, and can decode high-level commands and provide decoded high-level commands to a hard application module. group. The data collection server further includes an application module control unit 720 that processes the higher commands provided by the TW2024PA 18 1303032 unit 710 in accordance with the higher order command to control the operation of the application module 740. In other words, the high-level command processing unit 710 determines the presence or absence of the corresponding application module 740 according to the high-level command provided by the application program interface (API) 700. If the corresponding application module 740 exists, the high-level command is decoded into an applicable module. The 740 recognizes the command and transmits the command to a corresponding mapping unit or control message. The application module control unit 720 includes mapping units 721, 723, 725, 727 temple port 729, and interfaces 722, 724, 726, 728 # port 730 for application modules 741, 743, 745, 747, and 749 °, respectively. / receiving processing module mapping unit 721 receives the data request message from the host and the analysis system, transmits standard data and/or non-standard data to the host or analysis system through the network, and the process device and the non-standard data detecting unit Communicate to collect standard and non-standard data. The transmit/receive processing module mapping unit 721 receives the higher command of the higher command processing unit 710, maps the high command into a device-level instruction that can be processed by the transmit/receive processing module 741, and transmits/transmits The receiving processing module interface 722 provides the above component command to the transmit/receive processing module 741. If desired, the transmit/receive processing module 741 can include a physical conversion interface to directly communicate with the process device and the host. In this example, the physical switching interface includes a communication switching interface. For example, a switch 〇 standard data collection module mapping unit 723 receives the data request message from the host and/or the analysis system. The memory system stores the above information request message, interprets the above information request message, and collects standard data from the process equipment. To achieve this goal, the standard data collection module mapping unit 723 receives the higher order command from the higher command processing unit 710 to collect non-standard data, the image high command into a component command that can be recognized by the standard data image module 723, and through standard data collection. The module interface 724 provides the above component commands to the standard data collection module 743. TW2024PA 19 1303032 The non-standard data collection module mapping unit 725 determines whether a message (which is not necessarily used to analyze the process device) is included in the data request message transmitted by the host and/or the analysis system. In addition, the non-standard data collection module image unit 725 also collects non-standard data from non-standard data detection units connected to the process equipment. To achieve this goal, the non-standard data collection module mapping unit 725 receives the high order command of the higher command processing unit 710 to collect non-standard data, and the image high command is a component command that can be recognized by the non-standard data collection module 745, and through the non-standard The standard data collection module interface 726 provides the above component commands to the non-standard data collection module 745. The data conversion module mapping unit 727 can convert the non-standard data collected by the non-standard data detecting unit into standard data or data that can be recognized by the host or the analysis system. To achieve this goal, the data conversion module mapping unit 727 receives a high level command that can convert non-standard data into standard data or data from the higher command processing unit 710. The image high command can be recognized by the data conversion module 747. The component command is provided, and the component command is provided to the data conversion module 747 through the data conversion module interface 728. When the data collected from the process equipment and the non-standard data detecting unit is transferred to the host or the analysis system, the filtering processing module mapping unit 729 selectively processes the collected data, allowing only the unneeded and requested data to be transmitted. To this end, the filtering process module mapping unit 729 receives a higher order command that selectively transmits the data that has been collected or generated by the higher command processing unit 710 to the host and/or analysis system. The image high command is a component command that can be recognized by the filter processing module 749, and the component command is provided to the filter processing module 749 through the filter processing module interface 730. For example, when the host or the analysis system requests block-based data from the process device and the non-standard data detection unit, the above architecture can remove unnecessary data from the data, or be selective. The collected or generated data is transferred to the host and/or analysis system, thus avoiding an increase in network load and the TW2024PA 20 1303032 avoiding a reduction in manufacturing line stability. As described above, the data collection system of the embodiment of the present invention is structured in such a manner that an independent data collection server is provided between the host and the process device to transmit a message between the host and the process device, and in addition, the standard is executed. Collection of data, collection of non-standard data, conversion of data, and synchronization of data. Therefore, according to the present invention, standard data and non-standard data can be collected in time without a large amount of time delay, so that the problem of data synchronization can be solved. Further, since the host requirements are not changed, the problem of the system burden can often be overcome in the conventional semiconductor process system. In addition, the present invention can also improve the network burden by filtering the data and the internal data-sync type configuration to control the transmission of unnecessary data to the host or the analysis system. - This problem often occurs in traditional semiconductor process systems - can be solved. Referring to Figure 8, the signal flow of the collection system is described below. Most semiconductor process equipment uses the Generic model for communications and control of Manufacturing equipment (GEM) communication protocol to describe the type and meaning of information between the host, the analysis system, and the process equipment. Therefore, the analysis system supporting the GEM communication protocol collects data according to the GEM communication protocol. When the process equipment does not support the GEM communication protocol or there is non-standard data, it is necessary to increase the process of supporting the GEM communication protocol. This embodiment of the invention provides a process flow table for the above scenario. In the data collection system of the embodiment of the present invention, the host and/or the analysis system collects the required information by the state variable IDentifier (SVID) through the data collection server according to the GEM protocol. Most Status Variable Identification Codes (SVIDs) are collected and managed by Report IDentifier (RPTID). Further, the host and/or analysis system defines and manages the TW2024PA 21 1303032 Collection Event IDentifier (CEID) manufacturing system < uses this code to notify the occurrence of an event, such as an emergency. In this case, the event identification usually refers to the non-host required identification code to define the occurrence of an event. For this purpose, the data collection system defines this message rule so that the message can be exchanged through the analysis system to exchange the above-executed functions. The GEM protocol defines some standard messages 'refer to Figure 8, which are related to the data in the data collection server. Figure 8 shows the signal flow of the data collection system according to the embodiment of the invention. Refer to S1F3 in Figure 8. And the S1F4 message, respectively, represents a host or analysis system for querying a selected state variable identification code (ISVID) transmitted from the process device, and a manufacturing message responding to the former message. For example, the temperature of the first operating room 1 of the process equipment is set such that the pressure of the SVID 1 and the second operating room 2 is set to SVID 2, when it is necessary to know the temperature of the first operating room 1 and the second operating room. When the pressure is 2, the host writes SVID 1 and SVID 4 ' in the message S1F3 and then transmits the message S1F3 to the process device, and the process device adds the temperature of the first operation room 1 and the pressure of the second operation room 2 to the message S1F4, and then Send the message S1F4 back to the host or analysis system. The messages S2F23 and S2F24 respectively represent messages that issue a tracking command to the process device and periodically report different status variable identification codes (SVIDs), and a message that the process is set to respond to the aforementioned message. The messages S2F23 and S2F24 can be used when it is difficult to collect the state variable identification code using the S1F3 and S1F4 messages. For example: The host commands the process device to detect the state variable identification code SVID 1, SVID 2, SVID 3, SVID 4 by the message S2F23, and the process device responds to the message S2F23 with the message S2F24. Messages S2F33 and S2F34 represent messages that use a bundle of SVIDs to redefine and delete the report identification code (RPTID), and messages that the process device responds to. TW2024PA 22 1303032 Messages 82?35 and 82?36 represent the connection report identification code package (&131111 (14 of RPTIDs) to the collection event identification code (CEID) or delete an existing connection message, and the process equipment The message responding to the above message. The messages S2F37 and S2F38 represent the message of enabling or disabling the event identification code (CEID) and the message of the process device in response to the message. In other words, when the process device reports an event The host may define a report identification code (RPTID) related to the collection event identification code (CEID) and a state variable identification code (SVID) related to the report identification code (RPTID). In this example, the event identification code (CEID) is collected. The code packets are transmitted by state variables, which are already present and combined with each other when the event occurs. Such a variable identification code packet is referred to as a report identification code (RPTID). The message S2F33 is used by the host to define a majority. The status variable identification code is a report identification code and a message for deleting the report identification code, and the message S2F34 refers to the message that the manufacturing system responds to the aforementioned message. Message S2F 35 is a message that connects a majority of the state variable identification code to the collection event identification code or does not connect the state variable identification code to the collection event identification code, and the message S2F36 refers to the message that the manufacturing system responds to the aforementioned message. The message S2F37 refers to whether the host is right or not. The specific collection event identification code is used to report to start or close the message of a specific event, and the message S2F38 refers to the message that the manufacturing system responds to the above message. The messages S6F2 and S6F1 represent the trigger state variable (SV) or the state variable identification code (SVID), respectively. The periodic report responds to the message of the tracking instruction and the report message generated by the manufacturing system in response to the aforementioned message. For example, the message S6F1 is a message used by the manufacturing system to respond to the host's tracking command to increase the value of the state variable identification code. The message S6F2 is used by the host to respond to the message of the process device message S6F1. The messages S6F12 and S6F11 respectively represent the message that causes the manufacturing system to report an event with the special TW2024PA 23 1303032 collection event identification code, and the manufacturing system generates the response to the aforementioned message. Report message. When process equipment needs When the host reports an event, it will use the S6F12 and S6F11. For example, in the example where the collection event identification code CEID i is defined as the program start and the collection event identification code cmD2 is defined as the completion of the card insertion, the process device transmits The message S6FU is written to the host that collects the event identification code, and the host transmits the S6F12 to the process device in response to the message S6fu. The messages S6F15 and S6F16 respectively represent the host to learn about the collection event identification, the report identification code, and the learned from The manufacturing system queries the message collecting the sorrow variable of the event identification code and the report message generated by the manufacturing system in response to the aforementioned message. In other words, the message S6F15 means that when the host wants to know the report identification code and the state variable identification code value for a specific collection event identification code, the host uses a message for collecting an event identification code and transmitting a caution. Messages 8矸16 are information used by the manufacturing system to write report identification code and status variable identification code values for the specific collection event identification code, and to communicate a note that responds to the aforementioned message. The messages S 6 F! 9 and s 6 F 2 0 respectively represent the message that the host uses to know the state variable identification code contained in the report identification code queried from the manufacturing system, and the message generated by the manufacturing system in response to the aforementioned message. In this example, the state variable represents the value of the state variable (four) code. Meanwhile, in the above message, 'S' and "F" respectively represent, stream, (stream) and "function" (function) for classifying messages exchanged between the host and the manufacturing system. For example, message S6F19 represents a message corresponding to stream 6 and function 19. Further, the above information is transmitted by the host through the data collection server to the data request message of the process device, and the collected data is transmitted from the process device to the host and/or the analysis system through the data collection server. / As described above, according to the non-standard data collection system of the present invention, the data collected by the data collection TW2024PA 24 1303032 servo benefits can be synchronized and processed so that the data is provided as provided by the process equipment. The following is a non-standard data collection method for the above non-standard data collection system. 9 through 12 are flow charts showing a method for collecting non-standard data in a semi-institutional system according to an embodiment of the present invention. Referring to Figure 9, first, the data collection server specifies a variable identification code (SVID) defined by the analysis system as non-standard data to be processed by the process device during the phase si〇 analysis, wherein The state variable identification code is determined by the analysis department: Thereafter, at stage S12, the data collection server receives a data request message from the host and/or the analysis system. In stage S14, the data collection server stores the data demand message sent from the host and/or the analysis system in the memory. Next, in stage S16, the data collection server determines whether the state variable identification code and the collection event identification code (CEID) are included in the data requirement message defined by the process device. The process equipment does not support (4) news, which is interpreted as the non-standard data requirements of the stage. Then in stage S2G, (4) of the non-standard f material is performed. At the same time: Stage S22, - From the data request message, the data collection server removes information that is not supported by the process equipment. Further, in step S24, the request message for non-standard data is transmitted to the process device. Referring to Fig. 10' at stage S26, the data collection server receives (4) standard materials via a non-fresh data unit connected to the process equipment. In this example, the methods described in Figures 5a and 5b are used to complete the collection of non-standard data. Thereafter, in stage S28, the data collection (four) server collects the reference data. At this stage, S30, when Zen Qunzi (4) (4) ^... Tian Wei 枓 枓 枓 ' ' ' ' ' ' ' ' 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料 资料Where the message is - and the messages contain the settings from the process when performing non-standard data collection
TW2024PA 25 1303032 備所接收_標準型式狀態變數制碼，此時所收集的非標準資 料會被處理成同步資料。 〜之後，在階段S32’所收集到的非標準資料被轉換成標準資 料。在此情形中，資料收集伺服器收集來自製程設備的標準資 料。此階段中，標準型式的資料包含主機或分析系統所能辨認的 資料，或是標準資料。 之後，在階段S34,參照儲存在記憶體的資訊，來修復原始 Λ心在卩自奴S36，一回應訊息係結合從製程設備收集的標準資 料、從非標帛資料偵測單元收集的非標準資料、以及根據修復訊 息轉換成的非標準資料。之後，在階段S38，上述回應訊息係傳 送給分析系統且/或主機。 另一方面，依據第11圖，如果由非標準資料偵測單元接收 到一負認知訊息(negative acknowledge; NAK)，其係回應 S2F33, 剛5與S2F37訊息中至少-個訊息，則資料收集舰器取消之 前在階段S40, S42與S44所執行的非標準資料收集動作。 更進一步，在此發明實施例的半導體製程系統中，非標準資 料收集方法係包含-通訊切換架構（c〇m_icati〇n細血叫 scheme)，其可防止當系統操作不正確或不穩定時因主機與製程 設備間的通訊干擾導致半導體製造線停擺。換言之，此發明的非 標準資料收集方法包含能將主機與製程設備間的通訊切換成直 接通訊的通訊切換架構，縱然半導體製程設備設計成穩定操 可能導致資料收㈣服器的不穩定，此架構在當中可提供通訊切 換。-利用通訊切換架構的繞道方法（bypassing meth〇d)描述於 下。 第12圖係繪示一通訊繞道方法的流程圖，此方法可應用於 本發明實施例的半導體製程設備非標準資料的收集方法。TW2024PA 25 1303032 Receiver _Standard type status variable code, the non-standard data collected at this time will be processed into synchronous data. After that, the non-standard materials collected in the stage S32' are converted into standard materials. In this case, the data collection server collects standard data from the process equipment. At this stage, the standard type of data contains information that can be recognized by the host or the analysis system, or standard data. Then, in stage S34, referring to the information stored in the memory, the original heart is repaired in the slave S36, and the response message is combined with the standard data collected from the process device and the non-standard collected from the non-standard data detecting unit. Information and non-standard data converted from repair messages. Thereafter, in stage S38, the response message is transmitted to the analysis system and/or the host. On the other hand, according to Fig. 11, if a negative acknowledgement (NAK) is received by the non-standard data detection unit, it responds to at least one message of S2F33, just 5 and S2F37 messages, and the data collection ship The device cancels the non-standard data collection actions performed in stages S40, S42 and S44. Further, in the semiconductor manufacturing system of the embodiment of the invention, the non-standard data collection method includes a communication switching architecture (c〇m_icati〇n blood called scheme), which can prevent the system from being operated incorrectly or unstablely. Communication interference between the host and the process equipment caused the semiconductor manufacturing line to stop. In other words, the non-standard data collection method of the invention includes a communication switching architecture capable of switching communication between the host and the processing device into direct communication, even though the semiconductor process device is designed to be stable, which may result in instability of the data receiving device. Communication switching is available in it. - The bypassing method (bypassing meth〇d) using the communication switching architecture is described below. Fig. 12 is a flow chart showing a method of communication bypassing, which is applicable to a method for collecting non-standard data of a semiconductor process equipment according to an embodiment of the present invention.
TW2024PA 26 1303032 參照第12圖，應用於本發明非標準資料收集方法之通訊繞 道方法，係包含當資料收集伺服器出現不正常情形時，在資料收 集伺服器與連接主機與製程設備間設置一看門狗（watchd〇幻機 制。若用更洋細的方式描述，本發明通訊繞道方法包含階段S44， 用於決定通過看門狗機制時電源是否關掉，階段S46決定通訊切 換是否開啟。當階段S44之決定結果為電源關掉，以及階段S46 之決疋結果為通訊切換開啟時，階段s6〇可執行主機與製程設備 間可直接互相通訊以繞道資料收集伺服器。 更進一步說明，本發明之通訊繞道方法，包含階段S48，若 電源未關閉且通訊切換尚未開啟時，其可用於更新看門狗機制。馨 此外，通訊繞道方法還包含階段S5〇，當繞道資料收集伺服器與 甚至看門狗機制關掉或暫停狀態（time〇ut state)時，其可執行通訊 切換，使主機與製程設備直接互相通訊。 · 另一方面，在上述的實施例中，本發明的資料收集系統允許 - 貝料收集伺服器執行部分主機與製程設備的功能，如此可有充足 的彈性以提供系統的運作。 進步，田偶發狀況發生時，系統的功能可以擴充延伸到配 備儲存媒體的電腦中，藉由上述收集方法，此媒體用以儲存上述翁 半導體製程系統之非標準資料收集方法的執行程式。 根據此發明，在不需要改變現存的半導體製程系統規劃與增 加負載下’標準資料與非標準資料可及時彼此同步。此外，根據 此^明’過去無法應用在現存的主機與製程設備的各種功能，此· 處均可有效地配置在主機與製程設備之間的資料收集伺服㈣ · k伸功此中。更進-步，本發明的優點在於可應用於現存的所有 設備與新設備。當根據此發明資料收集伺服器與配合分析系統， 應用於半導體製造線時，半導體工㈣生產值將可觀地增加。TW2024PA 26 1303032 Referring to FIG. 12, the communication bypass method applied to the non-standard data collection method of the present invention includes setting a view between the data collection server and the connection host and the processing device when the data collection server is abnormal. The dog (watchd illusion mechanism. If described in a more subtle way, the communication bypass method of the present invention includes a stage S44 for determining whether the power is turned off when the watchdog mechanism is passed, and the stage S46 determines whether the communication switch is turned on. The result of the decision of S44 is that the power is turned off, and the result of the step S46 is that the communication switching is turned on, the stage s6〇 can directly communicate with each other between the host and the processing device to bypass the data collecting server. Further, the present invention The communication bypass method includes stage S48, which can be used to update the watchdog mechanism if the power is not turned off and the communication switch has not been turned on. In addition, the communication bypass method also includes stage S5〇, when the bypass data collection server and even the gatekeeper When the dog mechanism is turned off or paused (time〇ut state), it can perform communication switching, making the main Direct communication with the process equipment. On the other hand, in the above embodiments, the data collection system of the present invention allows the -bee collection server to perform functions of a part of the host and the process equipment, so that there is sufficient flexibility to provide the system. Operation. Progress, when the incident occurs, the function of the system can be extended to the computer equipped with the storage medium. The above collection method is used to store the execution program of the non-standard data collection method of the above-mentioned Weng semiconductor manufacturing system. According to the invention, the standard data and the non-standard data can be synchronized with each other in time without changing the existing semiconductor process system planning and increasing load. In addition, according to this, the past can not be applied to various existing host and process equipment. Function, this can effectively configure the data collection servo between the host and the process device. (4) k. In this case, the advantage of the present invention is that it can be applied to all existing devices and new devices. When the server and the matching analysis system are collected according to the invention, applied to semiconductor manufacturing At the time of the line, the production value of the semiconductor workers (4) will increase considerably.
TW2024PA 27 1303032 綜上所述，雖然本發明已以一較佳實施例揭露如上，然其並 非用以限定本發明，任何熟習此技藝者，在不脫離本發明之精神 和範圍内，當可作各種之更動與潤飾，因此本發明之保護範圍當 視後附之申請專利範圍所界定者為準。 TW2024PA 28 1303032 【圖式簡單說明】 第1圖係繪示習知半導體製造系統之結構示意圖。 第2圖係繪示習知ΕΑΡ資料同步型（data_sync type)半導體 製造系統之結構示意圖。 第3圖係繪示習知管理（agent)資料同步型半導體製造系統 之結構示意圖。 第4圖係繪示本發明一種收集半導體製造系統資料之系統 架構示意圖。 第5a和5b圖係繪示根據第4圖資料收集系統的資料同步方 法之示意圖。 第6圖係繪示根據第4圖資料收集系統的資料收集伺服器之 示意圖。 第7圖係繪示根據第6圖資料收集伺服器的模組結構示意 圖。 第8圖係繪示根據此發明實施例之資料收集系統的訊號流。 第9圖至第11圖係繪示根據本發明實施例之半導體製造系 統收集非標準資料之方法流程圖。 第12圖係繪示根據本發明實施例之半導體製造系統收集非 標準資料之通訊繞道方法流程圖。 【主要元件符號說明】 110 :製程設備 120 :電腦主機 420 :資料收集伺服器 421 ··記憶體系統 422 :中央處理單元（CPU) TW2024PA 29 1303032 423 ： 高速主記憶體 424 ： 算術邏輯單元 425 ： 辅助記憶體 426 ： 登錄器 427 ： 輸入系統 428 ： 控制器 429 ： 通訊糸統 430 ： 主機 440 ： 分析系統 700 ： 應用程式介面（API) 710 ： 高等命令處理單元 720 ： 應用模組控制單元 721 ： 傳送/接收處理模組映像單元 722 ： 傳送/接收處理模組介面 723 ： 標準資料收集模組映像單元 724 ： 標準資料收集模組介面 725 ： 非標準資料收集模組映像單元 726 ： 非標準資料收集模組介面 727 ： 資料轉換模組映像單元 728 ： 資料轉換模組介面 729 ： 過濾處理模組映像單元 730 ： 過濾處理模組介面 740 ： 應用模組 741 ： 傳送/接收處理模組 743 ·· 標準資料收集模組 745 ： 非標準資料收集模組 30TW2024PA 27 1303032 In summary, the present invention has been described above in terms of a preferred embodiment, and is not intended to limit the invention. Various modifications and refinements are intended to be included in the scope of the invention as defined by the appended claims. TW2024PA 28 1303032 [Simple Description of the Drawings] Fig. 1 is a schematic view showing the structure of a conventional semiconductor manufacturing system. Fig. 2 is a block diagram showing the structure of a conventional data_sync type semiconductor manufacturing system. Fig. 3 is a block diagram showing the structure of a conventional data synchronization type semiconductor manufacturing system. Figure 4 is a schematic diagram showing the architecture of a system for collecting semiconductor manufacturing systems according to the present invention. Figures 5a and 5b are schematic diagrams showing the data synchronization method of the data collection system according to Fig. 4. Figure 6 is a schematic diagram showing the data collection server of the data collection system according to Figure 4. Fig. 7 is a schematic view showing the structure of a module for collecting a server according to Fig. 6. Figure 8 is a diagram showing the signal flow of a data collection system in accordance with an embodiment of the present invention. 9 through 11 are flow charts showing a method of collecting non-standard data by a semiconductor manufacturing system in accordance with an embodiment of the present invention. Figure 12 is a flow chart showing a method of communication bypass for collecting non-standard data in a semiconductor manufacturing system according to an embodiment of the present invention. [Main component symbol description] 110 : Process device 120 : Computer host 420 : Data collection server 421 · Memory system 422 : Central processing unit (CPU) TW2024PA 29 1303032 423 : High-speed main memory 424 : Arithmetic logic unit 425 : Auxiliary Memory 426: Logger 427: Input System 428: Controller 429: Communication System 430: Host 440: Analysis System 700: Application Interface (API) 710: Advanced Command Processing Unit 720: Application Module Control Unit 721: Transmit/Receive Processing Module Image Unit 722: Transmit/Receive Processing Module Interface 723: Standard Data Collection Module Mapping Unit 724: Standard Data Collection Module Interface 725: Non-standard Data Collection Module Mapping Unit 726: Non-standard Data Collection Module interface 727: Data conversion module image unit 728: Data conversion module interface 729: Filter processing module image unit 730: Filter processing module interface 740: Application module 741: Transmission/reception processing module 743 ·· Standard Data Collection Module 745 : Non-standard assets Collection module 30
TW2024PA 1303032 747 :資料轉換模組 749 ··過濾處理模組 S10 :指定狀態變數辨認碼（SVID)為需要的非標準資料 S12 :從主機接收資料需求訊息 S14 :儲存在記憶體中 S16 :決定是否已定義狀態變數辨別碼（SVID)與收集事件辨 認碼（CEID) S18 :解釋未定義的非標準資料訊息 S20 :收集非標準資料 S22 ·•移除未定義的訊息 S24 :傳送訊息至製程設備 S26 :收集非標準資料 S28 :接收非標準資料 S30 :處理非標準資料成同步的非標準資料 S32 :轉換非標準資料成標準資料的型式 S34:參照儲存在記憶體的資訊來修復原始訊息 S36 :根據修復訊息結合從製程設備收集的標準資料與非標 準資料（架構回應訊息） S38 :傳送回應訊息給分析系統和主機 S40 ··接收回應S2F33，S2F35，S2F37訊息之訊息 S42 :接收到負認知訊息（NAK)否？ S44 :電源是否關閉？ S46 :通訊切換是否開啟？ S48 :更新看門狗機制（watchdog) S50 :看門狗機制是否關掉或暫停狀態（timeout state) ? S60 :通訊切換（繞道） TW2024PA 31TW2024PA 1303032 747: Data conversion module 749 ··Filter processing module S10: Specify the status variable identification code (SVID) as the required non-standard data S12: Receive the data request message from the host S14: Store in the memory S16: Decide whether Defined State Variable Discrimination (SVID) and Collected Event Identification Code (CEID) S18: Interpret undefined non-standard data message S20: Collect non-standard data S22 · Remove undefined message S24: Send message to process device S26 : Collecting non-standard data S28: Receiving non-standard data S30: Processing non-standard data into synchronized non-standard data S32: Converting non-standard data into standard data type S34: Refer to the information stored in the memory to repair the original message S36: According to The repair message combines the standard data and the non-standard data collected from the process equipment (architectural response message). S38: Send the response message to the analysis system and the host S40. · Receive the response S2F33, S2F35, S2F37 message S42: Receive negative cognitive message ( NAK) No? S44: Is the power supply turned off? S46: Is the communication switching enabled? S48: Update watchdog mechanism (watchdog) S50: Is the watchdog mechanism turned off or paused (timeout state)? S60: Communication switching (bypass) TW2024PA 31