An evolutionary multi-objective framework for business process optimisation
Graphical abstract
Highlights
► First research attempt aimed at addressing the challenges of business process optimisation using evolutionary multi-objective optimisation algorithms. ► Introduces a representation technique that models the visual and quantitative elements of a business process design. ► Proposes a framework that integrates the representation scheme with the state-of-the-art EMOAs. ► Demonstrates the efficacy and the utility of the proposed framework for real-life business process optimisation.
Introduction
A business process is perceived as a collective set of tasks that when properly connected and sequenced perform a business operation. The aim of a business process is to perform a business operation, i.e. any service-related operation that produces value to the organisation. In response to increasingly volatile and competitive environments, organisations are examining how their core business processes may be re-designed to improve business performance and market responsiveness. The design and management of business processes is a key factor for companies to effectively compete in today's volatile business environment. By focusing on the optimisation and continuous improvement of business processes, organisations can establish a solid competitive advantage by reducing cost, improving quality and efficiency, and enabling adaptation to changing requirements. It has been argued in [23] that for systematic and holistic business process planning based on business process automation, there must be techniques that support modelling, analysis and optimisation of business processes.
The focus of this paper is on business process optimisation and analysis and not on modelling [1], [12]. Business process optimisation is a new area that provides a formal methodology for improving business processes based on specific objectives. To achieve formal improvement however, business processes also require a formal representation. Despite the abundance of business process modelling techniques in literature, only a few are capable of capturing a business process in a quantitative way so that it can be optimised.
The aim of this paper is to investigate the application of evolutionary multi-objective optimisation to the new domain of business process optimisation. The rationale for adopting a multi-objective optimisation approach towards business processes lies in the following:
- 1.
Business process optimisation is inherently a multi-objective optimisation problem due to the variety of factors that a business process can be evaluated with. Dealing with multiple objectives can make this research more appealing and applicable to real-life business process optimisation problems.
- 2.
Evaluation business processes based on a series of relevant factors ensures that this research is versatile in dealing with different objectives for different business goals at a time. The capability of simultaneously addressing a series of customised quantifiable objectives ensures the generality of the research and its potential applicability in a wider context of business process improvement initiatives.
- 3.
On the contrary, a single-objective optimisation framework focuses on a particular objective (e.g. cost reduction) and thus loosing its generality and its advantages over context-specific business process improvement approaches that target a specific aspect of a business process (e.g. Six Sigma).
This paper presents the first research attempt aimed at addressing the challenges of business process optimisation using evolutionary multi-objective optimisation algorithms. It introduces a business process representation approach and an evolutionary multi-objective optimisation algorithm that constructs optimised business process designs based on specific process requirements. The utilisation of evolutionary techniques provides the advantage of working with a population of designs, thus providing the capability of generating a set of diverse optimised business process designs in the presence of multiple objectives.
Section snippets
Related work
It is commonly accepted that a holistic approach towards business processes should capture a business process (business process modelling), provide the necessary means for bottleneck identification and performance analysis and – finally – generate alternative improved business processes based on specified objectives. But often the last part (business process optimisation) is overlooked – if not completely neglected [7]. It is argued in [21] that business process modelling is essential for the
Representation and composition of business process designs
This section introduces the proposed representation of a business process designs and also an algorithm that is able to compose business processes based on specific requirements.
The business process optimisation problem
This section presents the formulation of the business process optimisation problem based on the proposed business process representation. The problem formulation based on the parameters shown in Fig. 1a assumes that there are more than one process attributes used as optimisation objectives and thus is considered as a multi-objective optimisation problem. The multi-objective problem formulation for business process optimisation is as follows:
For a business process design with a set of nd tasks
Business process optimisation framework (bpoF)
The proposed business process optimisation framework (bpoF) applies a series of existing EMOAs to a business process design captured using the proposed representation. The aim of the framework is to fully utilise the proposed representation technique and the capabilities of the EMOAs in order to generate a series of alternative optimised designs. Based on the aim, the main operation of the framework is the generation and optimisation of business process designs. Fig. 4a demonstrates the inputs
Validation using real-life business process scenarios
During the development of this work, the proposed framework was evaluated by generating experimental business process scenarios. While the details can be found in [19], it must be noted that these experimental scenarios were generated in a manner that they could resemble real business process problems in terms of problem requirements – composing and optimising a business process using a library of tasks given the input and output process requirements.
This section validates the capability of the
Conclusions
The aim of this paper was to investigate the application of evolutionary multi-objective optimisation to the new domain of business process optimisation. This was achieved by: (i) representing business processes in a quantitative way, (ii) algorithmically composing business process designs based on specific requirements and (iii) identifying the optimal processes utilising the state-of-the art EMOAs.
This paper introduced a representation technique that models the visual and quantitative
References (25)
Business process modelling: review and framework
International Journal of Production Economics
(2004)- et al.
Staff scheduling and rostering: a review of applications, methods and models
European Journal of Operational Research
(2004) - et al.
Business process intelligence
Computers in Industry
(2004) - et al.
Algorithmical approaches to business process design
Computers & Operations Research
(2001) - et al.
Evolutionary algorithm for advanced process planning and scheduling in a multi-plant
Computers and Industrial Engineering
(2005) - et al.
A decision support system for business process planning
European Journal of Operational Research
(2000) - et al.
A comprehensive and automated approach to intelligent business processes execution analysis
Distributed and Parallel Databases
(2004) - et al.
PESA-II: region-based selection in evolutionary multi-objective optimisation
- et al.
A fast and elitist multi-objective genetic algorithm: NSGA-II
IEEE Transactions on Evolutionary Computation
(2002) - et al.
An improved dimension-sweep algorithm for the hypervolume indicator
Essential Business Process Modelling
The Pareto archived evolution strategy: a new baseline algorithm for multiobjective optimisation
Cited by (26)
Optimization of extended business processes in digital supply chains using mathematical programming
2021, Computers and Chemical EngineeringCitation Excerpt :Niedermann et al. (2011) formulate an optimization framework based on graph modeling and data mining to identify and recommend structural changes in process designs that can improve process performance, given a desired goal function. Other studies apply evolutionary algorithms to automate the design and redesign of business processes (Afflerbach et al., 2017; Mahammed and Benslimane, 2017; Vergidis et al., 2012). In the area of process intelligence and automation, industrial solutions exist to apply process mining to identify process steps that can most significantly improve performance if automated via robotic process automation (RPA) (van der Aalst et al., 2018).
Industry 4.0 and the circular economy: Resource melioration in logistics
2020, Resources PolicyCitation Excerpt :This discipline of business process management includes activities such as process modelling, automation, deployment, and optimization (Fig. 1). Each component supports superior results and operational excellence (Vergidis et al., 2007, 2008, 2012). Technology and business process management can support smoother mining operations in environments characterized by volatility, uncertainty, complexity, and ambiguity (Dehghani and Ataee-pour, 2012; Vom Brocke et al., 2014).
Performance Assessment of Business Process Optimization Algorithms Using a Prototype Dataset Generator
2024, International Journal of Information Technology and Decision MakingPareto-Optimal Trace Generation from Declarative Process Models
2024, Lecture Notes in Business Information ProcessingMulti-objective product allocation model in warehouses
2021, Techniques, Tools and Methodologies Applied to Quality Assurance in Manufacturing