[an error occurred while processing the directive]
RSS
Логотип
Баннер в шапке 1
Баннер в шапке 2
2015/02/16 13:13:10

Process Mining Process analytics

Process Mining is a technology of continuous recovery of cards of real business processes on the basis of the data which are stored in enterprise information systems. The constructed process maps are used for calculation of the expanded analytics allowing to increase operational efficiency of the enterprise, profitability of its products or services.

Content


The term Process Mining was entered into use of Vil van der Aalst, by professor of the Eindhoven technical university and Queensland technical university who is considered the ideologist of this technology. Principle of Process Mining very simple: if business process is executed in this or that information system, then on the basis of "traces" of its execution it is possible to recover the actual algorithm of process in the form convenient for the subsequent analysis[1].

Process Mining methods

  • the automated research of processes (extraction of models of processes of the event log)
  • compliance check (i.e. control of deviations by comparison of models and magazines)
  • the automated creation of simulation models,
  • expansion of model, recovery of model,
  • forecast of statuses
  • based on historical data, recommendations.




  • Continuous collecting and data processing
  • Automatic creation of model on the basis of information traces
  • The forecast of the choice of the scenario or time of end for yet not terminated processes
  • The uniform tool for data collection from all jobs of the company
  • Analysis of all chain of events, but not separate steps
  • Control of deviations from regulations and compliance to standards
  • Real-time data analysis and support of implementation of organizational changes


The Process mining methods are capable to take knowledge from the event logs usually available in modern information systems

What the process analytics is applied to

Implementation of tools of process analytics opens a number of opportunities, for example[2]:

  • Arrange internal processes. Simple example: large company in about tens of branches. In each of them the same process can proceed differently and occupy a different amount of time. Use of the software tool process mining will allow to understand as far as correspond (and whether correspond in general) processes regulated what risks for business are born.
  • Make management decisions on the basis of reliable data. At the heart of technology — use of the logs received from information systems. It is impossible to influence them, they cannot be hidden or forged therefore influence of a human factor is completely excluded.
  • Save. At liquidation or increase in capacity of "bottlenecks". In Eneco, having implemented the similar tool, announced saving 15 million euros. We, working with the large telecommunication operator within a pilot project, defined the amount of 75 million rubles — the company could save so much, robotizing transaction of attachment of files.
  • Raise a customer loyalty. Thanks to the deep analysis it is possible to prevent possible reputational and financial risks at early stages that considerably will raise a customer loyalty.
  • It is clear to visualize the actual process. Process mining is technology not to look how systems interact with each other and to find errors in it, and to look how employees and different departments are connected with each other. The manager, having studied the card of process, will be able to understand how he proceeds in reality where there are bottlenecks why these or those changes do not result in desirable result on what it is necessary to pay special attention.

Advantages and shortcomings of Process Mining

Advantages

  • The automated recovery of a business process model based on log files
  • Possibility of business process analysis to the level of a separate copy (a negative way)
  • Opportunity the facts to prove to the management inefficiency of the existing business processes on the basis of the facts
  • Possibility of regular monitoring of processes

Shortcomings

  • Lack of necessary detail of log files in information systems
  • Complexity of interpretation of data in information systems
  • Problems with correctness of data in information systems
  • Lack of the approved analysis technique of the recovered business process

Tasks for the Process Mining tools

Key task of Process Mining – recovery and the analysis of the actual business process. In this case the main thing that process was fully automated and was interesting to management for the purpose of the subsequent analysis. Key advantages of Process Mining in front of the existing tools of the "manual" analysis of business processes: leaving from traditional methods of the description through interviewing of representatives of business and manual process modeling by the business analyst in the form of graphic model for benefit of objective assessment of the occurring reality on the basis of the data which are in information systems.

"For management the Process Mining tools give a fine opportunity of condition monitoring of business processes in real time on the basis of actual data, but not the possible "pseudo-reports" collected in the manual mode", – Anastasia Martynenko, the senior consultant of BMP Logic company notes. An opportunity to rely on the facts at decision-making is one of key advantages of Process Mining.
"We use Process Mining for determination and identification of real processes of the organization", – Bondarenko Aleksandra, the chief consultant of QPR Software notes. Carrying out regular estimates and efficiency analysis of processes and also visualization of the course of processes is also implemented by the company in these tools.

Now it is still difficult to select industry specifics of application of Process Mining, but it is already clear that the technology will be used for interaction with a large number of clients. "In such areas as health care, transport, tourism, the banking sector and insurance, projects were already carried out and the researches which showed efficiency of Process Mining are executed, – Vladimir Rubin, the leading researcher of the International scientific and educational laboratory of the process focused information systems of Higher School of Economics National Research University notes. – Development processes of the software, job analysis of users, assessment of software performance, the analysis of quality of service, quality of a support service and work of contact center – all this the tasks solved by Process Mining".
Already there are Russian companies which work with Process Mining. For example, the Russian company ICL Services uses several years these tools in the work. "We used Process Mining by optimization and reengineering of the processes and also for identification of the processes automated, but not described, – notes, Yeremeyeva Aygul, the Head of service of organizational development ICL Services. – Also Process Mining is applied by us as the instrument of express diagnostics of the problems arising in processes of provision of services of IT outsourcing and in audits of internal business processes".

The Process Mining tools simplify quality control of work of employees in respect of compliance to regulations. "The back of a medal are a lack of a possibility of creation and adjustment or change of regulations directly in the tool", – Anastasia Martynenko notes. Most of clients, according to her, after the first acquaintance to technology ask a question: "And whether it is possible to draw something here? And how it is connected with a modeling tool of processes?".

Differences of Process Mining from Data Mining

  • Data mining is mainly used for search of hierarchical dependences in large volumes of data. For example, in what channels what categories of clients what types of goods buy and how often.
  • On an input tables with diverse data from different domains move.
  • Uses multidimensional views (cubes) with a possibility of level variation of detailing (different levels of aggregation) of information.

  • Process mining concentrates not on semantic interrelations of data, and on data view in the form of processes.
  • On an input transaction data on accounting items move. Usually act as such objects (Tasks, Orders, Requests, Dresses and so on). Event logs, auditor traces, data on events and statuses of objects are an example of transaction data (whether it be an object status or change of responsible division).
  • Uses methods of sampling of data for creation of a process model on the most representative scenarios in process. Process mining looks for not just communications between data: its task consists in defining communications between steps of process, deviations from normal process, factors of influence of a naotkloneniye, efficiency of process, a process stsenarnost and also bottlenecks in process.

Examples of projects of use of Process Mining

Ruukki (metallurgical company). Customer service process is analyzed, data were loaded from two different systems: SAP and Salesforce that allowed to consider them in uniform process. Shortcomings under the authority of data on clients of CRM were revealed that significantly slowed down passing of the order. Correction of the revealed shortcomings using the Process Mining tools – QPR Process Analyser – allowed to accelerate customer service and to increase the general service quality.

Caverion (engineering services and services for the industry, solutions for the real estate and industrial enterprises in Northern and Central Europe). Using QPR Process Analyser measurement of efficiency of basic processes of the company was taken. On the basis of the data obtained from the ERP system the reason of delays when passing processes was established: slow preparation of accounting documents upon accomplishment of services. As a result process performance was increased. The company made the decision on regular use of the Process Mining tool for efficiency evaluation of real business processes.

Vaisala (production of the metallurgical equipment). Using QPR Process Analyser processes of technical support, sales, deliveries, services in repair and marketing processes were analyzed. Data were collected from the systems of Salesforce and Oracle BI. Thanks to the made analysis of processes the company could change and optimize some of the processes and also reduce operational cost by means of quick response to changes in the course of processes.

Obstacles for Process Mining

However, despite all progressiveness tekhnologiiprocess Mining, mass application it for the present did not find. The pacing restraining factors at the moment are the low awareness of business on this method of the analysis and improvement of processes and also need of significant investments into development of own competences allowing to apply this method. "There is a scepticism concerning need of acquisition of the next specialized software and offers of the Process Mining tools as services in the market still small amount", – Aygul Yeremeyeva notes.

In addition to organizational factors there are also technology obstacles. "Insufficient process automation does not allow to prepare data for Process Mining and unfair use by employees of the existing information systems, for example untimely data entry or replacement of information, does implementation difficult, – marks out Alexander Bondarenko. – Data preparation and their extraction from the implemented business applications can be in such cases more expensive, than traditional process of obtaining information on business processes by means of an interview with the staff of the company".

The shortage of specialists and ignorance of opportunities of Process Mining and also heterogeneous data sources in the company and their chaotic status are obstacles for implementation of tools, Vladimir Rubin notes. He also considers that implementation of such systems is prevented by fear of transparency of results of work and data security about internal processes.

Data sources for Process Mining

The high level of process automation – a key factor for Process Mining use, and the different systems can be sources here. "Of course, the best sources are the systems which architecture was initially projected for support of processes and management of flows of works, – notes, Aygul Yeremeyeva. – These are the systems of the class ERP/EAM, CRM BPMS ITSM. They store historical data on state change of an object, whether it be an asset, work assignment or the address of the user".

Similar opinion and at Aleksandra Bondarenko. "ERP-, CRM- SCM- systems, support systems of users and processing of incidents can become data sources for Process Mining", – she says.

Actually information for the analysis of process can be collected from any transaction system. The following classes of systems can be sources for Process Mining: Data Warehouses, OLAP based repositories, ERP (Enterprise Resource Planning), PDM (Product Data Management), CRM (Customer Relationship Management), DMS (Document Management Systems), Trouble Ticketing, Service Management Systems (ITSM, etc), SCM (Software Configuration Management), APM (Application Performance Management), Bug Tracking, Project Tracking. "The main thing that in a system there was a set of event logs and traces on the basis of which the actual process is recovered", – Vladimir Rubin comments.

Anastasia Martynenko generalizes: "The main criterion of determination of whether a system source is suitable for Process Mining - it is the answer to very simple question: Whether "It is possible to recover data on process course of execution from the saved-up data?"" Analyzing results of use of the Process Mining tools, it is possible to see that the technology actively develops as a result of its approbation on practical cases in business companies. There is a confidence that with increase in level of automation of the existing business processes the number of users of Process Mining will only grow, increase in operational efficiency is a trend of the last years.

Process Mining software products

• QPR Process Analyzer (PA) • Celonis Process Mining • Proceset Process mining

You See Also

Notes