智慧自動化是一系列從流程自動化到人工智慧的策略技術概述, 結合機器人流程自動化 (機器人程式自動化) 具有光學字元辨識功能 (光學字元辨識), 智慧字元識別 (ICR), 流程挖掘, 深度學習, 機器學習, 認知技術, 自然語言處理, 語音辨識合成, 機器視覺, 以及許多其他 人工智慧 (人工智慧) 創造能夠思考的端到端業務流程, 學習並適應端到端業務流程. 也涵蓋了從流程發現的整個流程, 和流程自動化，透過自動化、持續的資料收集和理解資料的含義，使用資料來管理和優化業務流程.
什麼是智慧自動化?
智慧自動化是指借助自動化技術和人工智慧來預測客戶和使用者的需求，讓企業更接近數位轉型的軌道. 因此, 機構及其工作人員能夠更專注於提供高價值的任務. 例如, 企業可以建立附著情感的業務關係，為客戶和員工提供更有趣的體驗.
智慧自動化展示數位轉型整體解決方案, 主要基於業務流程管理，包括根據業務需求隨時協調使用者以及任務和系統的機器人流程自動化. 另外, 它利用分析和人工智慧(特別是機器學習) 發展自動化和智慧決策，實現端到端的成功管理.
最後, 值得注意的是，這一趨勢的另一個關鍵特徵在於企業所使用的不同系統的集成. 整合有助於防止系統資料重複出現. 更重要的是, 用戶只需在一個平台上工作.
智慧自動化的工作原理
智慧自動化混雜多種關鍵技術, 其中包括流程自動化, 機器學習, 業務流程自動化, 智慧型路由器和自然語言處理. 所有這些技術都是相互關聯並廣泛應用的. 也, 它們是數位轉型大型計劃的一部分.
智慧捕捉
智慧採集能夠自發性識別並提取文件中的信息, 紙本或數位形式. 它是光學字元辨識的混合體, 智慧字元識別, 光學標記識別以及條碼識別. 透過使用混合物, 可以捕獲結構或非結構數據.
工作流程自動化
工作流程自動化 是一種軟體, 它定義了一系列任務，並根據相關流程自動執行這些任務，無需人工幹預. 構成大型業務流程管理解決方案的一部分. 此類解決方案有助於優化業務流程並提供即時可見性，以便做出明智的決策並提供更好的業務服務.
機器學習
機器學習是人工智慧的一部分, 系統是從數據中學習的, 模式和演繹. 這使得系統能夠以最少的人工幹預來應對緊急情況.
機器人流程自動化
隨著機器人的加工自動化, 組織能夠在重複的應用程式和系統上執行任務. 能夠比較快速方便地融入現有的IT系統架構中, 這帶來了直接的好處，例如減少手動任務, 消除資料輸入錯誤並減少週轉時間.
智慧自動化使用了哪些技術?
它由相互關聯實施的四項重要技術組成, 因此，讓我們將它們分解並解釋每種技術所扮演的角色:
業務流程管理（BPM）
業務流程管理作為流程自動化技術發揮作用, 其中包括人員的有效協調, 系統和數據. BPM的目的是確保業務和營運流程的基礎結構穩固. 因此, 它可以被視為一個組織的基礎. 自動化複雜處理需要人中斷資料輸入與決策. 該系統將在特定時刻用於控制操作和數據生成以及計算和數據集成等數據存儲.
機器人流程自動化(機器人程式自動化)
Robotic process automation refers to a technology that aims to decrease manual intervention in the computer program.
Through the major interaction with “high-grade” applications, robotic process automation can perform work. These applications are relative to machine languages and program codes. In simple words, It is a kind of software that can simulate the interactive coordination between humans and normal computer applications.
人工智慧
Artificial intelligence stands for the devices which resemble the intelligence of people. 換句話說, it is the discipline that attempts to create a machine system that can achieve learning and deduction like humans.
Artificial Intelligence is comprised of various elements such as machine learning, 深度學習, 自然語言處理, visual recognition, big data and so on.
一體化
One of the biggest pain points facing enterprises lies in the connection and integration of systems since each system or software contains its own traits.
They usually provide an interactive application programming interface (應用程式介面) that is typically based on some standards such as SOAP (applied in Web services) and REST (based on the HTTP protocol).
智慧自動化的優勢是什麼?
Business values can be enhanced by lowering costs and boosting speed accuracy as well as transparency. Automation directed by machine learning, deep learnings, computer visions, text analysis as well as Chatbots is going to exert the greatest influence on the operation process such as user communication management, data capturing, business procedure management and case management.
透過最大限度地減少重複性的手動任務並在適當的時候消除瓶頸，可以增強員工的體驗，讓員工更專注於更高價值的工作. 轉向更具策略性並專注於建立關係 (科技和機器不能做什麼), 員工在整個數位轉型過程中發揮前所未有的重要作用.
合適的使用者可以在正確的時間接收系統提供的正確訊息，提升使用者體驗 (以及所有利害關係人的經驗). 使用智慧自動化使組織能夠透過重塑業務和推動新的交付成果來思考競爭，從而提升客戶體驗.
智慧自動化的真實案例有哪些?
With the implementation of robotic process automation in various industries, the scenarios covered by RPA have extended from rule-based, simple tasks automation to complicated process automation that is in need of Artificial Intelligence capabilities.
Compared with traditional RPA, AI-enabled RPA has more advantages in reading unstructured data, helping to make decisions, ensuring the accuracy of performing tasks and bridging human-robot interaction tasks.
金融場景: 收據處理
Commodities sold in stores require customers to pay at the mall payment counter and return the second link to the store after completing the payment. The mall and the store will periodically settle the payment through the card records from the POS machines. 在這個過程中, stores have to reconcile the sales records in the system with the actual payment amount (IE. the number of small bills), which includes a lot of data and a lot of manual reconciliation work.
RPA plus AI can realize automatic processing of financial receipts. UiBot Mage recognizes the electronic version of small bills, extracts vital information such as the date and sales volume, and reconciles the recognized information with sales records in the system.
Even if a small receipt contains sales records for multiple items (例如. six small receipts corresponding to ten sale records), the RPA, combined with AI robot, can perform reconciliation in an accurate way.
Manually, only a very small number of unaligned accounts need to be handled, significantly decreasing the manual reconciliation workload.
法律服務場景: 合約資訊的提取
The contract text is typical unstructured data, and generally requires manual entry of structured information into the contract management system. Even with the traditional “OCR recognition plus information extraction”, the average recognition accuracy is only around 75%. 更重要的是, a large amount of rules have to be written to collect the information.
With the combination of RPA and AI, it can intelligently extract crucial information such as a contract name, a party A, a party B, subject matter and quantity.
A company’s legal department has to deal with 3 contracts with completely different templates and contents (plain text, text plus table, text plus image, horizontal and vertical versions), which are time-consuming and laborious to handle manually.
UiBot Mage conducts analysis of the picture content with the help of document and form recognition to automatically extract the elements of contract information, providing an out-of-the-box solution for the contract management procedure, which not only lowers corporate labor costs but also decreases the risk caused by human operation errors.
銀行場景: 用戶銀行帳戶的處理表格
A bank processes 3,000 manual account opening forms from users every day, requiring employees to manually enter the form data 3,000 次.
With the wide application of RPA in the banking field, 500 forms can be handled automatically through the Robotic Process Automation. 然而, another 2,500 forms need to be processed manually because the manual writing finished by users is not fully standardized.
With the addition of Artificial Intelligence capabilities, the Robotic Process Automation can dramatically raise their understanding and are able to automatically handle over 2,800 of the 3,000 account opening forms, which is not only more accurate but more efficient. 所以, fewer than one-tenth of the rest of forms needed to be processed manually.
金融場景: 了解您的客戶程序
KYC (Know Your Customer) is an important compliance process in the financial industry. Banks, insurance, securities and other institutions need to collect customer data, verify customer identity and background, and validate customer information to fully understand their customers.
Most of this customer information comes from different documents and different system platforms. It takes at least a week for employees to manually process the data for verification.
機器人程式自動化, combined with AI, automatically identifies and reads data, collects customer information across systems, and verifies and reviews customer information with pre-defined parameters to filter out suspected fraudsters. Employees can spend only a few minutes completing the verification and validation process, which not only enhances processing efficiency but also saves a lot of time.
政務場景: 在積分視窗處理高頻事務
RPA plus AI is an important technical means to help connect the system platforms of different government departments, open up data sharing, make the business process more convenient and efficient, and empower the digital transformation of the government.
Take the service matter of “sewage discharge into the drainage network permit issuance filing” as an example: the entire business process requires a large amount of information and low fault tolerance rates, which requires more than 60 repetitive operations per day. 所以, it is time-consuming and labor-intensive.
With the help of RPA plus AI, the window staff only need to enter the credit code of the applicant units. Then the robots will automatically search the required information to complete the entry and print, and automatically scan and sort and file after the staff approved the seal. The whole process can be finished in two minutes, and the speed is constant with low error rates.
電力場景: the process of smart electricity
In the planning of each power grid company’s vigorous construction of intelligent electricity consumption process, RPA plus AI is applied to electricity consumption application data review and entry, electricity consumption data monitoring, electricity bill collection, automatic processing of electricity consumption complaints, intelligent customer service of electricity consumption and other process links to realize scientific statistics of the whole electricity consumption process, so that end terminal users can use electricity conveniently and quickly.
