An algorithmic method for business process reengineering and service automation in the public sector

Digital transformation of public services offers significant opportunities for improving the delivery of services to citizens, businesses, and other stakeholders in society. However, implementing digital transformation initiatives can be challenging, particularly in the public sector where processes and systems can be complex and bureaucratic, and as there no clear step-by-step guidelines to achieve a systematic, rigorous and repeatable result. The present paper aims at offering a valuable contribution to the field of digital transformation in the public sector, providing a practical and effective algorithmic method for business process reengineering and service digitalisation. The presented algorithmic approach to support the digital transformation of public sector services utilises process modelling tools, design patterns, innovative heuristics, simulation software and decision support systems. The method is demonstrated through a case study which involved the redesign of several hundreds of both internal processes and outward looking services in Greek local administration. Application results show significant improvements in process efficiency and effectiveness, repeatability and increased stakeholder satisfaction.


INTRODUCTION
The public sector is responsible for providing a wide range of services to citizens and businesses, ranging from healthcare to education, transportation, support of business activities, justice, defence and public safety.In recent years, there has been a growing emphasis on improving the efficiency and effectiveness of these services, while also ensuring that they are responsive to the needs of stakeholders.Service reengineering is a process that involves redesigning and optimizing existing services to improve their quality, efficiency, and effectiveness.The goal of service reengineering is to enhance the overall performance of the service, while also reducing costs and improving customer satisfaction.
Service reengineering has become increasingly important for the public sector, as governments seek to deliver more services with limited resources.The need for service reengineering is particularly acute in the context of aging populations, increasing demand for healthcare services, and changing citizen expectations.Thus, service automation in public sector has been a common target for governments around the world, as they seek to improve service delivery and reduce costs.Service automation refers to the use of technology to automate tasks and processes that were previously performed manually by human employees.According to a report by Andersen et al. [1], governments can benefit from automation in a number of areas, including citizen services, back-office operations, and regulatory and compliance functions.
The purpose of this paper is to explore the concept of service reengineering and automation in the public sector and provide a novel algorithmic method, by means of a repeatable, adaptable set of transformation steps that could be applied in a variety of processes and services.
The paper is structured as follows: In section 2, an overview of the key drivers and challenges associated with service reengineering in the public sector is presented, together with an overview of methods and tools that are used.The section concludes in stating the key shortcomings in current methods, which formulate the need for the proposed method.Section 3 presents the goals, the key steps and the supporting tools for the proposed algorithmic reengineering method.In section 4, the application of the method in the case of a large service reengineering project in Greece is presented, involving several hundreds of services that were modelled, reengineered, simulated and discussed with the stakeholders.Finally, section 5 presents the benefits, the limitations and the future possibilities of the overall proposed approach.

STATE OF THE ART 2.1 Business Process and Service Reengineering
Business process reengineering (BPR) is the radical redesign of business processes to achieve dramatic improvements in critical measures of performance such as cost, quality, service and speed.BPR is a comprehensive approach to improving business processes that involves questioning existing processes, analysing them, and redesigning them to achieve improvements in performance.The goal is to achieve significant and sustainable improvements in organisational performance by fundamentally changing the way work is done.
According to Hammer and Champy [2], the term "reengineering" means starting over with a clean sheet of paper and rethinking the way work is done.They define BPR as "the fundamental rethinking and radical redesign of business processes to achieve dramatic improvements in critical, contemporary measures of performance, such as cost, quality, service, and speed." BPR involves a fundamental rethinking of business processes, with a focus on improving overall performance rather than simply automating existing processes.
Davenport [3] describes BPR as "the radical redesign of business processes to achieve dramatic improvements in organizational performance."He identifies four key principles of BPR: (1) the focus on business processes rather than individual tasks, (2) the reorganization of work around business processes rather than functional specialties, (3) the use of information technology to enable new ways of working, and (4) the involvement of all stakeholders in the design of new processes.
During the last 20 years, BPR and digital transformation have been very closely related, as both involve the fundamental rethinking and redesign of business processes to achieve better organisational performance.Digital transformation involves the use of digital technologies to transform business operations and create new value for customers, while BPR focuses on rethinking and redesigning business processes to achieve significant and sustainable improvements in organizational performance.In today's digital age, BPR often involves the adoption of digital technologies to enable new ways of working, improve process efficiency, and enhance customer experience.According to a study by the Bughin, J et l.[4], companies that fully digitize their operations can achieve up to 5 times more revenue growth than their less digitized counterparts.Furthermore, BPR can serve as a catalyst for digital transformation by identifying areas where digital technologies can be leveraged to achieve better business outcomes.By rethinking existing processes and identifying opportunities for digital transformation, companies can achieve significant improvements in performance and create a competitive advantage in the digital marketplace.
Reviewing the state of the art in BPR and service reengineering, reveals a multiplicity of approaches, methods, tools, ideas and frameworks, that can be summarized as following: • Digital transformation of business processes involves using standard digital technologies to improve the efficiency, effectiveness, and agility of a process [5], resulting in benefits such as cost savings, increased productivity, and better customer experiences [6].
• Cloud computing can provide the necessary infrastructure and platforms for digital transformation projects [7] while the use of APIs (Application Programming Interfaces) can enable greater integration and collaboration between different digital technologies used in a business process [8].• Business process management (BPM) tools can be used to model, analyse, and optimize business processes as part of digital transformation efforts [9] while process mining techniques can be used to analyse process data and identify opportunities for improvement as part of digital transformation projects [10] and simulation techniques can help organizations to test and optimize their processes before implementing them in the real world [11].• Cybersecurity is an important consideration in digital transformation projects, and organizations need to ensure that their processes are secure and compliant with data protection regulations [12].• Advanced digital technologies can be used to transform business processes, such as artificial intelligence [13], blockchain [14], Internet of Things [15], while the use of chatbots and virtual assistants can improve customer service and support in a business process [16] and Robotic Process Automation (RPA) can be used to automate repetitive manual tasks in a business process [17].Also, data analytics and machine learning can be used to improve the accuracy and effectiveness of business processes [18], while Digital Twins can be used to create virtual models of a process that can be used for simulation, testing, and optimization [19].
Further to the above mainstream of methods, tools and technologies that drive business processes and services reengineering, literature also contains a vast collection of additional guidelines and soft methodologies for tackling strategic, organizational and cultural aspects, such as: • The success of digital transformation projects depends on effective change management, stakeholder engagement, and leadership [20].• Design thinking methodologies can be used to ensure that digital transformation projects are user-centric and focus on solving real customer problems [21].• The use of agile methodologies can help organizations to iterate and adapt their digital transformation efforts more quickly [22].• Digital transformation projects need to be aligned with the overall strategy and goals of the organisation to ensure that they deliver value and support business growth [23].• Digital transformation projects require a culture of innovation and experimentation, with a willingness to try new technologies and approaches in order to achieve success [24].

The lack of a Systematic Approach for Service Reengineering
Although several methodological frameworks, methodologies, guidelines and tools exist for business process and service reengineering, the lack of a repeatable, standardized, systematic, thus algorithmic method for achieving the same result when transforming a process from paper-based to fully automated and digital is lacking.This gap has already been identified both by governments and industry: In Europe, the lack of a systematic approach to service reengineering in the public sector has been identified as a major obstacle to improving service delivery [25,36].Despite some successful initiatives, there is still a need for a standardized methodology that can be applied across different countries and sectors.
In the United States, the lack of a systematic approach to service reengineering has been attributed to the fragmented nature of the public sector [26].There are thousands of different government agencies and departments, each with their own processes and procedures, which can make it difficult to develop a standardized methodology for service reengineering.
In China, the lack of a systematic approach to service reengineering in the public sector has been linked to the country's complex bureaucratic structure [27].Government agencies in China are often hierarchical and siloed, which can make it difficult to identify and prioritize areas for improvement.
The lack of a systematic approach to service reengineering is a challenge that is widely recognized by international consulting firms: • McKinsey & Co notes that many public sector organisations lack a standardized methodology for identifying and improving inefficient processes [28].• Deloitte also highlights the need for a systematic approach, noting that the public sector often struggles with fragmentation and silos, which can make it difficult to identify areas for improvement and coordinate efforts across different agencies and departments [29].• Accenture notes that a lack of resources can also be a barrier to service reengineering in the public sector.Many organizations are operating with limited budgets, and may not have the resources to invest in new technologies or process improvements [30].• PwC points out that the lack of a systematic approach to service reengineering can also result in a lack of clarity and consistency in service delivery.Without a standardized methodology, different departments or agencies may have different approaches to delivering the same services, leading to confusion and inefficiencies [31].• EY notes that resistance to change can be a significant challenge to service reengineering in the public sector.Government employees may be hesitant to adopt new technologies or processes, and there may be a lack of incentives to drive change [32].

PROPOSED METHODOLOGY AND TOOLS 3.1 Principles
After analysing several methodologies and existing frameworks for business process and service reengineering in the public sector [37], a step-by-step methodological framework was developed, comprising tools, modelling and reengineering processes and guidelines, available semantic components and other supporting elements.The key principles of this framework are presented in the following paragraphs.

Service Automation Patterns
Concerning service automation, the 5-level framework that is used by the European Commission has been adopted, which defines the following service automation and digitisation levels [33]: • Level 1: The service is just described on the internet.
• Level 2: Additionally to the service description, samples of forms and documents are provided online.• Level 3: The service recipient can apply for the service online, but the outcome is to be produced manually.• Level 4: The service is provided fully online, in one stop and without human intervention.• Level 5: The service is provided proactively, and is personalised.
The proposed framework aims at delivering the highest possible level of automation for each service or process to be transformed.
The above automation levels lead to the creation of relevant generic Service Patterns, that is generic illustrations of the service flow, which are then taken in mind when performing the reengineering of each service.

Service Value
The creation of final value by a public service is pending to its ability to create or lead to the creation of one of the following results for citizens or businesses [33]: • Registration in a base registry, change or deletion of this registration (e.g., civil registry, business registry, land registry, etc.) • Financial result for the state or the citizen / business (e.g., payment of tax or levy, payment of allowance, payment of due amount, etc.) • Granting of a license to the citizens or the company for some activity supervised by the state, within the framework of its regulatory role (e.g., building license, license to establish a business), • Production of a tangible or intangible result (e.g., implementation of a technical project, delivery of goods, waste collection, finding a job, etc.) • Creation of a necessary personal document (e.g., passport, identity card, driver's license, health business license, etc.) to the extent that this document has not already been fully digitized

Service Categories, based on the value they create
In order to focus the transformation upon services that create final value for the citizens or businesses and not digitize trivial processes or services [38] that exist only because of the uncomplete digitization of the state (e.g., a paper certificate that should be available online), the proposed method identifies three service categories: • Category A: Services that will remain.This category contains value-generating services (services that generate one or more of the 5 tangible results, as presented above) and so they have to be reengineered and automated to the maximum extent.
• Category B: Services that will eventually be phased out.This category includes services that do not create real value [38], but are needed by other services, due to lack of interoperability in the public sector.Such services may be transformed to lower automation levels, as they will sooner or later be abolished.
• Category C: Services that should not be transformed.
This category includes services that can be merged into other services, are trivial (not subject to monitoring or optimisation) or otherwise should be abolished.
Service Archetypes Through analysing thousands of services offered by the public sector to citizens and businesses [34], the following 12 Service Archetypes have been created, which correspond to abstract flows for each different service type: As with the Service Automation Patterns, for each Service Archetype a generic process flow is created, to act as a generic guide for the service transformation.

Tools and Resources
The proposed service reengineering method is supported by a set of tools and resources, aiming at the creation of a sustainable and manageable infrastructure that can support the BPR processes in long term.The needed infrastructure, as shown in Figure 1, includes the following elements: • A Business Process Modelling Notation (BPMN) Editing Tool, able to create relevant process descriptions.Such a tool should not be just a diagram editor but offer an accessible database of all BPMN elements (processes, documents, process steps, decision making rules, organizational entities, etc).• A Process Simulation Toolkit, connected to the BPMN Tool, able to simulate the behaviour of each process under various demand levels and with various public sector staffing levels.• A Process / Service Metadata Repository, holding all the supportive definitions and information about the modelled processes and services, as well as Codelists and other Controlled Vocabularies that are important for the precise definition of service details (e.g., nationalities, currencies, citizen groups, etc).Such metadata are typically compatible with definitions such as the Core Public Service Vocabulary Application Profile (CPSV-AP) widely used in the European Union [35].• A Repository of existing public sector Infrastructures, containing the existing or future information systems that play a role in the provision of the services to be reengineered.• eXtensible Markup Language (XML) Editing Tools, for the design of messages to substitute the existing documents, in case data transformation is also part of the reengineering project.

• The Service Archetypes and the Service Automation Patterns
Repositories, typically in the selected BPMN Tool, containing the abstract flows of the corresponding standard patterns and archetypes.• A Legal Database, which contains the various laws, decisions, decrees, and other legal and statutory system elements that control the provision of the services under reengineering.This database will then depict the various changes that will be needed in the legal system when the new services come into operation.• A Common Database, where all the textual, diagrammatic or other elements are stored.• A Consultation and Data Gathering Subsystem, which is essential for getting input from the public sector officials, both during the initial data gathering phase, but also during the various consultations and approvals for the new proposed services.
Additional to the above, several simpler tools and resources might be used depending on the needs of each reengineering project, such as spreadsheets, text documents or presentation material.

The Modelling and Reengineering Algorithm
Utilising the tools presented in Section 3.2, the proposed method comprises of twelve distinct steps to be followed by the reengineering team.These steps, as depicted in Figure 2, belong to three Step 1: Gather Process DataWhere the project team gathers all needed information for the services to be reengineered.This might include existing process descriptions or diagrams, service metadata, codelists, descriptions of information systems and more.In case no modelling or specialized description tools are in place, and no descriptions exist, information is to be gathered through specifically designed questionnaires and even structured interviews with the public sector officials that know the processes to be modelled.
Step 2: Model Process in BPMN and Process RepositoryIn this step, the "as-is" status of the processes to be modelled is inserted in the BPMN tool and supportive information is populating the other repositories (metadata, codelists, controlled vocabularies, legal elements).The next steps 3 to 9 are then performed for each specific process / service.These steps can also run in parallel, speeding up the overall reengineering task.
Step 3: Decide on Process CategoryFor each process to be modelled and reengineered, a decision should be made on the category of the process, as described in Section 3.1.Services that belong to Category A will undergo a full reengineering, service of Category C will not be affected (as the will soon be phased out or abandoned) while services of Category B will not be fully automated, to cater for the interim period until final phase out.
Step 4: Select Process ArchetypeIn this step, each process is assigned to its corresponding archetype.Through this assignment, the modeler already gets an overview of how the final, reengineered process should look like.For example, all payments to the public sector should be 8 made through the designated payment banking systems and gateways, all new entries to Base Registries should follow the Once-Only principle and only ask for the essential citizen data, etc.
Step 5: Decide on Process Automation LevelAt this step, and based on the process archetype but also taking into account the existing or future public sector infrastructures, the target automation level for each process to be reengineered is set.Typically, most processes should target levels 5 (personalized, proactive provision) or 4 (full, one-stop automated provision).In cases where some manual process is needed (e.g.site inspection for the provision of a permit) the automation level can be lower.
Step 6: Modify Process Input Stage This step contains the following sub-steps and modelling principles and decisions: • All services will be implemented via a suitable Internet Information System (IIS) of the public sector or a suitable mobile device application (MDA).If such a system does not exist, it has to be assumed that will come into place.• If the service is reorganized at levels 3 or 4, the initiation of the service is made by the citizen or the business, who chooses to get it via the appropriate IIS or MDA.In case of a level 5 service, the service must be initiated by the operator (with an event generated by the IIS and a message sent to the citizen/business to complete its execution.• In case a proof of identity of the applicant for the service (passport, identity card, etc.) is needed, it should be replaced with a form of digital authentication, usually provided by the central authentication gateway of the state or by a third party (e.g., bank or other digital identity authorized provider).• In case the service applicant must provide original information in written form (e.g., a signed statement or application), this will be substituted by a digital form to be created in the respective IIS.• In case a public document is needed for the provision of the service (e.g., a certificate), this will be replaced by a web service call from the IIS to the appropriate Base Registry.If such a registry does not exist, it has to be assumed that will come into place (e.g., a certificate of residence is substituted by a web service call to the respective Residence Base Registry).• In the event that a required document does not fall into the above categories, the modeler considers whether it can be replaced with an appropriate online service.Otherwise, the document should be provided to the public sector IIS in a digital, non-machine-editable format (scanned).Of course, this will lead to lowering the targeted automation level.
Step 7: Modify Process Flow This step aims at modifying all the process steps, between input and output, and contains the following sub-steps: • Eliminate all the steps of the process that concern the circulation of information in the form of a document internally in the organization such as reading emails, typing, sending, or sharing documents, receiving documents from another department, etc.This information will be automatically processed by the IIS of the organization, which will appropriately notify the necessary recipients (typical Workflow Management System functionality) • Replace all the checks and decisions made by human operators by automatic processing whenever the legal aspect accepts this change, provided that all input elements are digital and mechanically processable (e.g., fields in digital forms, or fields in digital messages of online services) and all decision-making rules can be followed by the IIS, as they are typical if-then-else structures.More complex, tacit or unstructured, decision-making the deployment of a more complex artificial intelligence system should be considered.• Replace any communications with third parties during the implementation of the service, by calling an appropriate online service of the third party, via web services.If the service of the third party does not exist, it must be assumed.• Unify all further communication, verification and record-keeping steps within the organization, which will now be automatically carried out by the IIS (e.g., notification of the management, creation of logs, co-signing of documents, etc) In the case of services with target level 4 or 5 (full automation) no human is allowed to interact with the process execution.Human intervention is only allowed in services with an automation level up to 3.
Step 8: Modify Process Output StageThis step involves the replacement of any type of paper or digital documents produced by the process, by: • Automated processes that inform the IIS of the organization.
• Messages sent via email or though mobile applications to recipient citizens or businesses.• Web Services that can be called by third parties when needed (e.g., when updating a base registry, no certificate has to be issued.The base registry will respond, when called via a web service, with the new information).• Calling a web service online services of third parties, for the continuation of the execution of a service (e.g., payment via a banking system) Step 9: Create the new Process Model in BPMN and Process RepositoryIn this step the new process model is created, after performing all the modification to the "as-is" process as analysed in steps 6,7 and 8.This includes creating the "to-be" BPMN model but also declaring all the new, assumed infrastructures and services in the appropriate repositories described in Section 3.1.
Step 10: Run Process Simulations and ConsultationProcess simulation is an important step for fine-tuning the newly designed services and processes.Such a step can be performed both after the reengineering of each service (individual simulation) but also including all the reengineered services, so that one can ger insights on the needed staffing and information systems capacity under real-life, complex scenarios.At this step, further fine-tuning can be made via online consultations with the public sector officials, when presented with the new services and the simulation results.An online consultation system is foreseen in Section 3.1, allowing for browsing the various new service flows and registering one's opinion or suggestions.
Step 11: Document needed Systems, Web Services and Mes-sagesAfter the completion of simulation experiments and consultations with public sector officials, the reengineered services can now be finalized with the creation of all the systems, web services and messages that have been assumed during the reengineering process.The deployment of such systems will be a crucial factor for the actual operation of the new services.
Step 12: Document new Organisational Units and StaffingThe final step of the method includes providing additional information on what will be the new non-systemic structures that will be needed for providing the services under redesign.This typically includes new organizational units and restructured existing organizational units, as well as staffing levels that can guarantee operation under the assumed demand, with the targeted response times.

APPLICATION
The presented algorithmic reengineering method was applied in a large-scale process reengineering project in Greece, that was implemented between 2018 and 2021, procured by the Greek Company for Local Development and Administration [34].The project involved the complete re-engineering of all internal processes and services offered to citizens and services by Municipalities and Regional Administrations in Greece.Each methodological step was carried out in absolute and continuous collaboration with the Municipalities and Regional Administrations as well as legal advisors.Such a vast project, which included the modelling and redesigning of thousands of different processes, was one of the factors that led to the creation of the presented standardized, tools-driven methodology.Table 1 provides an overview of the project size, in terms of processes / services to be reengineered.The project was implemented by a mixed team of public servants, process modelers, tools operators and developers that at times exceeded 100 members, with the authors of the present paper holding the core roles in project methodology, tools support and managing the modelling and reengineering process.Specific tools deployed included the ADONIS©commercially available BPMN and process management tool, around which a number of extensions were built, to allow for service metadata, controlled vocabularies, legal elements and XML management.Also, due to the lack of valid descriptions of processes within municipalities and regions, additional tools were developed for gathering initial data and documents through online collaboration with local government officials.
A sample of the process repository is given in Figures 3 and  4, where the "as-is" and the "to-be" BPMN models of a municipal service for enrolling foreigners in the Base Civil Registry is shown, The BPMN model shows the result of the core methodological steps (modification of project input, flow and output), where all documents and human tasks have been substituted by information systems and machine-operated, automated services, The project achieved its core targets, completely redesigning more than 1,000 services and internal processes in municipalities and regions, leading to a projected reduction of more than 80% in execution times of more than 64% of staff in working hours, and almost completely eliminating paper or digital documents from such processes, as shown in Table 2.

CONCLUSIONS
Service reengineering in the public sector can be an extremely demanding task, both because of the vast number of services offered   towards citizens and businesses but also due to the lack of standardized, step-by-step methodologies for achieving a repeatable reengineering result.This paper presented a repeatable, adaptable algorithmic method, consisting of a set of transformation steps that can be applied in a variety of processes and services.In essence, while the Algorithmic Method for Business Process Re-engineering and Service Automation in the Public Sector can serve as a guiding framework, its successful implementation requires careful consideration of each country's unique attributes and needs.Flexibility, adaptability, and a commitment to tailoring the approach will be key factors in realizing the benefits of automation across diverse public sectors.The method, which includes the applied principles, the needed information management tools and the exact transformation steps was applied in a large service transformation project for Greek Local Administration, comprising more than 1,300 internal processes and services to citizen and businesses, achieving very high reduction of service execution time, paper generated through the process and needs for staffing, as most of the services were made fully automated.
A shortcoming of the proposed algorithmic methodology relates to the very high level of targeted service automation, which in order to be fulfilled requires the development of a series of information systems and web services, which at the design time are nevertheless properly assumed and documented.
Such a method has the potential to be fully automated, so that process and service re-engineering for the public sector might become -itself -a fully automated, machine-drive process.

•
Registration, correction of data or deletion from a base registry • Grant of financial aid or allowance • Grant or extension of a license • Payment to the public sector • Collection by the public sector • Issuance of a certificate or permit • Issuance or replacement of civil or business documents • Appeal against a public sector decision • Provision of a participative service • Request for information by a citizen or business • Approval of a project outcome by the public sector • Submission of additional information to the public sector, for a service or process

Figure 1 :
Figure 1: The tools and resources of the proposed BPR method

Figure 2 :
Figure 2: Overview of the proposed BPR method steps

Table 1 : 383 Figure 3 :
Figure 3: The "as-is" status of a municipal service, in BPMN

Figure 4 :
Figure 4: The "to-be" status of a municipal service, in BPMN

Table 2 :
Greek Local Administration project achievements Element Value Total Number of Internal Processes Re-engineered 338 Total Number of Services Re-engineered 710 Number of new Base Registries Designed 186 Number of new Web Services / API's Designed 198 Overall average % reduction of average execution time 82% Overall average % reduction of staff for service execution 64% Overall average % reduction of paper 99%