Editor's recommendation:
With the acceleration of the digital transformation process of enterprises, a large number of information systems have been launched, but the problems of process operation management have gradually emerged. In order to improve the process operation capability of the enterprise, AsiaInfo cooperated with a provincial operator to launch an intelligent process governance operation mode, and tried to deeply integrate task mining technology and process mining technology based on RPA (Robot Process Automation) to realize process data-driven process changes The closed-loop iteration helps enterprises achieve the goal of "transforming growth drivers, improving process governance efficiency, and upgrading operating models". This article introduces the ideas and goals of the RPA automation process governance program, the specific steps of business process optimization, summarizes the characteristics and effects of the program, and looks forward to the future development of automated process governance. It is hoped that this article can provide a reference for enterprises concerned about the improvement of RPA technology and process automation capabilities.
01
Research background
With the release of the "Overall Layout Plan for Digital China Construction" (hereinafter referred to as "Digital China Plan") in 2023, the digital transformation of enterprises ushered in a positive institutional environment and clear policy guidelines. The country has vigorously promoted the development of digitalization, which has accelerated the pace of digital transformation of enterprises, organizations and the whole society. According to incomplete statistics, more than 95% of enterprises have carried out in-depth and comprehensive digitalization practices to varying degrees after the digital transformation has been fully launched in all industries. Developed a clear strategic plan for digital transformation. The digital transformation of domestic enterprises has become a new normal, and major enterprises are continuing to increase investment in digital transformation funds.
In practical applications, it is found that in the past, enterprises built a large number of information systems, and achieved rapid business expansion and replication through solidified business processes and standardized management. With the continuous development of the business and the constant change of the organization, a large number of information systems have been launched, and the problems of process operation management have gradually emerged:
● Invisible process path: The process lacks visual monitoring, and business personnel cannot grasp the actual running process of the process. Once stuck or blocked in the process, the process execution efficiency will be greatly reduced. Based on the many problems in the process, there is an urgent need for some way to realize process visualization, thereby improving the efficiency of process execution.
● Uncontrollable process details: At present, there are many office systems in enterprises, such as ERP, CRM, OA, etc. At the same time, problems such as large amount of data, difficult integration and analysis, time-consuming, inefficient and inaccurate manual analysis also arise. In addition, due to the complexity of the process, the location of the process bottleneck is not clear, so it is difficult to accurately locate the process stuck point. How to accurately discover problems and bottlenecks in complex business processes, so as to provide accurate analysis results and suggestions is a problem that needs to be explored.
● Improving process efficiency is not feasible: Once the process bottlenecks and stuck points in the process are accurately located, it is urgent to complete process optimization by automated means or provide process optimization suggestions based on a complete solution knowledge base, so as to Quickly solve practical problems such as process optimization and efficiency improvement.
02
Solution Exploration
In order to solve the above problems, enterprises and various solution providers have invested a lot of human resources to cooperate with the collaboration process to sort out, but the actual results are not satisfactory. The perspective of human combing results is not comprehensive enough, and the analysis is not precise enough, which cannot be used as an effective input for subsequent process optimization and efficiency improvement.
In order to assist enterprises to achieve the goal of organizational digital transformation, AsiaInfo tries to deeply integrate Task Mining and Process Mining based on RPA (Robotic process automation), in which the process mining technology is visualized Restore the business process trend, accurately find problems and bottlenecks, and use task mining technology to deeply analyze business operations and discover potential automation opportunities. RPA performs targeted process optimization based on the problems and opportunities found, so that the results of RPA process automation can also be directly used for optimization. And reconstruct the process to realize the closed-loop iteration of process change driven by process data. Help enterprises achieve the goal of "transforming growth drivers, improving process governance efficiency, and upgrading operating models".
03Practical
plan
In order to improve the process operation capability of the enterprise, AsiaInfo cooperated with a provincial operator to launch an intelligent process management operation mode.
(1) The overall idea and objectives of the program
This solution aims to introduce process mining technology, relying on AsiaInfo's AISWare AIRPA products to achieve sustainable enhanced process governance, facilitate process change, collect user operation behavior through task mining desktop recording, and use screen understanding to identify business processes. The front-end business event logs, combined with the back-end data of the business, form an integrated front-end and back-end process mining data.
This solution will complete the four major functional points of process visualization, bottleneck analysis, process design and process optimization:
● Process visualization: use the process discovery algorithm to visually reproduce the business process of the enterprise, that is, process the event log, generate flow chart data, calculate process indicators and display them in the foreground. It provides timely insight into the process, can truly restore the actual execution path of the business process, does not rely on manual research, and can understand the process execution in real time, objectively and continuously.
● Bottleneck analysis: Provides the ability to diagnose process problems in real time and discover process bottlenecks. It can quickly locate bottleneck node paths that affect process operation efficiency and analyze the causes. Realize joint drill-down analysis of process dimension and business dimension, and assist managers to analyze key KPIs of business processes from a process perspective.
● Process design: According to the process of bottleneck analysis and process insight, discover the blocking points and bottlenecks of the process, intelligently identify the breakpoints, blocking points, and stuck points in the process, and give suggestions for business optimization and reshaping, so as to promote the process Intelligent and automated transformation.
● Process optimization: By optimizing the bottlenecks and deviations, the actual process can be optimized, the existing business process can be improved, the process fluency can be improved, and the business transformation efficiency can be improved. For example, resources and personnel can be allocated at reasonable nodes, and detected Any abnormal link, so as to find out where the problem lies. Automated means can also be used to intelligently assess the degree of automation of process steps, thereby using RPA robots to achieve more efficient automation.
(2) Program details
Since every action that occurs in a software system, whether performed by a human or a robot, produces a record, these records become event logs. Based on the system log events that can be recorded in sequence, each event refers to an activity and is associated with a specific business scenario. After summarizing the additional information in all event logs, a "flow chart" form of the actual process can be formed. view information. After the process is visualized, use process mining technology to gain insight into the ability of the enterprise process, determine the gap between the ideal state and the actual state, and quickly find out the root causes of inefficiency and execution deviation. Finally, based on the analyzed results, a deeper level of excavation is carried out to complete the optimization of the business process.
Figure 1: Schematic diagram of an RPA-based automated process governance solution
By connecting with the internal and external IT systems of the enterprise, event logs are obtained from the system, and process mining algorithms are called to complete the process of data cleaning. After conversion, filtering, and cleaning, process event logs that can be used for integration are generated. After the event log is processed, it will be displayed visually to restore the real business scenario. Based on the restored business scenarios, follow-up intelligent analysis and intelligent prediction are carried out. Analyze the bottlenecks in the process through KPI analysis, filter setting and other process analysis tools, and give suggestions for optimization. Specifically, it includes the following parts:
Step1: Business log collection
Obtain log data from business systems. These logs will contain event IDs, time stamps, event names, user names, system information, etc., and are important data sources for process mining. Then the log data needs to be preprocessed to filter invalid and error messages.
Step2: Business log mapping
Based on the collected business log data, time information, event information, customer information, etc. are associated with the corresponding background log events, forming an integrated front-end and back-end process mining data, such as associating customer numbers, product information, etc. with the corresponding background log events. This can provide richer business background information for process mining and get more accurate analysis results. Specifically include:
● Event correlation: In each event of the system business log, add a field of business data and record the corresponding business information. For example, in the "select customer" event, add the "customer number" and "customer name" fields to store the selected customer information. This makes it easy to correlate business data with each specific log event.
● Case association: establish a Case ID for each system business log case, add the same primary key field in the business database, and store all business data information related to the case. In this way, all business data related to a case can be quickly found through the primary key.
● Time interval association: Based on the timestamp of the system business log, search for records in the business database within the same time interval, and extract these records as the corresponding business data of the log.
● Rule association: According to business rules, such as the combination of fields such as "customer number-product number-region" must be unique, to find the business data records that match the system business log events.
Step3: Build a business process model and visualize it
In this program, the genetic mining algorithm and simulated annealing algorithm are combined to mine the process Petri net model, DFG model, BPMN model, etc. from the log data. Combined use of multiple models can provide a comprehensive view of the process. At the same time, generate flow chart data, calculate process indicators and display them in the foreground. Provides timely insight into the process, can truly restore the actual execution path of the business process, does not rely on manual research, can understand the process execution in real time, objectively and continuously, and can clearly display the process steps and the logical relationship between each step.
Figure 2: Schematic diagram of business process visualization
Step4: Process compliance check
Process analysis: compare the real trend of the restored business process with the preset process standard and verify the compliance of process nodes and paths, and find out deviations in the process by analyzing business data, such as time-consuming, repeated rework, frequent Interruption, etc., and then promote the subsequent further optimization analysis.
Bottleneck analysis: Provides instant diagnosis capabilities for process problems and discovery capabilities for process bottlenecks, and can quickly locate bottleneck node paths that affect process operation efficiency and perform cause analysis. Realize joint drill-down analysis of process dimension and business dimension, and assist managers to analyze key KPIs of business processes from a process perspective.
According to the process of process analysis and bottleneck analysis, the blocking points and bottlenecks of the process are found, and the breakpoints, blocking points, and stuck points in the process are intelligently identified.
Figure 3: Schematic diagram of abnormal process error message
Figure 4: Schematic diagram of abnormal process statistics and analysis
Step5: Task Mining Analysis
Through process analysis and bottleneck analysis, specific problems are located, and then the in-depth analysis from the business level to the operation level is realized through task mining. Specifically pull through the front-end and back-end business data, accurately locate specific problems, push forward suggestions for follow-up business optimization and reshaping, and complete intelligent and automated processes.
Figure 5: Schematic diagram of task mining analysis
Step6: Process optimization
According to the problems identified by process mining and task mining, improvement measures are put forward. Use RPA to quickly implement process changes and run log records again to verify optimization results. This constitutes a closed loop of RPA-based process mining.
Figure 5: Schematic diagram of process optimization solution
(3) Program characteristics
● Front-end and back-end data integration, accurate identification of process bottlenecks: Support joint analysis of business data and desktop operation data, not only to discover business bottlenecks from the overall level, but also to accurately view specific business details through desktop operation data, so as to identify automation opportunities.
● Abundant process value-added tools to help enterprises understand business processes: provide rich process value-added tools, such as business KPI analysis, compliance inspection, bottleneck inspection, root cause analysis, continuous business monitoring, etc.
● Multi-modal mining algorithm to improve the accuracy of process mining: it provides different optimized mining algorithm support for different business scenarios, and can adaptively select the optimal mining algorithm according to different business scenarios.
● Multi-platform mass data collection and analysis, providing diversified data samples: supports Windows, Linux, Android, graphical interface and other system data collection, can collect various systems and types of data, and quickly perform extraction and correlation analysis.
04
Effect summary
This solution can be applied to the automation of work order operation process to speed up work order operation. In the work order processing process, there are situations where the process takes too long to run. By visualizing the business process, mastering the actual operation effect of the work order, and locating the specific process stuck point, it was found that the description, title and other information on the work order page took too long to fill in the entire process, and the processing measures The input of information and solutions takes a long time, and there is invalid content input, which leads to the failure or withdrawal of the work order approval. After locating specific problems, provide suggestions for process optimization. The specific results are as follows:
● Save manpower: It can save a lot of human resource investment, increase efficiency by 60% in a very short period of time, save labor costs by 40%, eliminate redundant or invalid business process branches, reduce IT system investment, and effectively improve business turnover Improvement meets the requirements of the enterprise management to improve efficiency and reduce costs.
● Agile value: Currently, 20+ business processes have been explored and analyzed, and 8 of them have been analyzed and optimized. Taking the collaborative cooperation process as an example, correcting 3 abnormal branches, shortening the process running time by 40%, reducing the page operation steps by 20%, greatly improving the efficiency of process governance, and effectively helping users achieve a comprehensive upgrade from process automation opportunity discovery to process governance optimization .
05
Future Outlook
AsiaInfo relies on the AISWare AIRPA platform to implement intelligent process governance to help industry users optimize organizational business processes, improve business response speed, and effectively improve organizational process management and operational efficiency.
Relying on the AISWare AIRPA platform, AsiaInfo adopts a one-click deployment method and has zero intrusion into the business system. It only needs to use the Task Mining recorder to collect the whole process of business personnel handling business and collect multiple system logs (such as business support system, operation support system, management support system, etc.) system), there is no need to modify the existing system, it can be used out of the box, it can be applied to various scenarios, and it is highly replicable and popular. It can be widely used in various industries, such as helping operators to strive for excellence in the construction of a new generation of digital intelligent operation and maintenance platform, group pilot work, digital intelligence capabilities, etc., and improve work shortcomings. Currently, Henan Mobile business process governance has been realized. Intelligent capabilities are improved, business processing time is reduced, customer satisfaction is improved, and the pace of digital transformation is accelerated.
At present, the importance of the development of the digital economy and the urgency of the digital upgrading of China's industries are self-evident, and it will become a new driving force for China's economic growth. In addition to policy encouragement and support, various enterprises and institutions are also deepening practice and exploration. According to a recent market survey, about 57.6% of organizational leaders currently list business process governance optimization as one of the important breakthrough capabilities for organizational development (recognizing the importance of processes for business development). Relying on the AISWare AIRPA platform, AsiaInfo actively responds to the development direction of the national digital economy, provides suggestions for the digital upgrading of Chinese industries, assists corporate process governance, improves corporate process transparency, empowers corporate process optimization, and helps companies better carry out digital practices.
References
[1] Process Mining：Data Science in Action.Van Der Aalst，2018.
[2] Detailed Explanation of RPA Technology: Foundation, Application and Future, AsiaInfo, 2022.
[3] Process Mining Handbook. Van Der Aalst,2022.