Business Process Management Forum

The complexity of manufacturing processes is increasing due to the production variety implied by mass customization of products. In this context, manufacturers strive to achieve flexibility in their operational processes. Business Process Management (BPM) can help integration, orchestration and automation of these manufacturing operations to reach this flexibility. BPMN is a promising notation for modeling and supporting the enactment of manufacturing processes. However, processes in the manufacturing domain include the flow of physical objects (materials and products) apart from information flow. Buffering, bundling and unbundling of physical objects are three commonly encountered patterns in manufacturing processes, which require fine-grained synchronization in the enactment of multiple process instances. Unfortunately, BPMN lacks strong support for this kind of dynamic synchronization as process instances are modeled and executed from a single, isolated point of view. This paper presents a mechanism based on BPMN 2.0 that enables process modelers to define synchronization points by using the concept of recipes. The recipe system uses a dynamic correlation scheme to control many-to-many interactions among process instances to implement required inter-instance synchronizations. We formally describe the involved BPMN patterns, implement and evaluate them in a manufacturing scenario in the high-tech media printing domain.

Declarative, constraint-based approaches have been proposed to model loosely-structured business processes, mediating between support and flexibility. A notable example is the Declare framework, equipped with a graphical declarative language whose semantics can be characterized with several logic-based formalisms. Up to now, the main problem hampering the use of Declare constraints in practice has been the difficulty of modeling them: Declare’s formal notation is difficult to understand for users without a background in temporal logic, whereas its graphical notation has been shown to be unintuitive. Therefore, in this work, we present and evaluate an analysis toolkit that aims at bypassing this issue by providing users with the possibility to model Declare constraints using their own way of expressing them. The toolkit contains a Declare modeler equipped with a speech recognition mechanism. It takes as input a vocal statement from the user and converts it into the closest (set of) Declare constraint(s). The constraints that can be modeled with the tool cover the entire Multi-Perspective extension of Declare (MP-Declare), which complements control-flow constraints with data and temporal perspectives. Although we focus on Declare, the work presented in this paper represents the first attempt to test the feasibility of speech recognition in business process modeling as a whole.

Mobile robots like drones or ground vehicles can be a valuable addition to emergency response teams, because they reduce the risk and the burden for human team members. However, the need to manage and coordinate human-robot team operations during ongoing missions adds an additional dimension to an already complex and stressful situation. BPM approaches can help to visualize and document the disaster response processes underlying a mission. In this paper, we show how data from a ground robot’s reconnaissance run can be used to provide process assistance to the officers. By automatically recognizing executed activities and structuring them as an ad-hoc process instance, we are able to document the executed process and provide real-time information about the mission status. The resulting mission progress process model can be used for additional services, such as officer training or mission documentation. Our approach is implemented as a prototype and demonstrated using data from an ongoing research project on rescue robotics.

Significant research work has been conducted in the area of process mining leading to mature solutions for discovering knowledge from structured process event logs analysis. Recently, there have been several initiatives to extend the scope of these analysis to consider heterogeneous and unstructured data sources. More precisely, email analysis has attracted much attention as emailing system is considered as one of the principal channel to support the execution of business processes (BP). However, given the unstructured nature of email logs data, process mining techniques could not be applied directly; thus it is necessary to generate structured event logs. Activities form a cornerstone component of BP that must be identified to obtain such structured logs. In this paper we propose to discover frequent activities from email logs. Existing approaches are usually supervised or require human intervention. In addition, they do not take into consideration activity business data (BD) perspective. In this paper, we introduce a pattern discovery based approach to tackle these limitations; we suggest mainly to discover frequent activities and their BD with unsupervised way. Additionally, our approach allows the detection of multiple activities per email and the automatic generation of their names while reducing human intervention. We validate our work using a public email dataset. We publicly provide our results to be a first step towards ensuring reproducibility in the studied area, which allows more practical analysis for further research.

Conformance checking describes process mining techniques used to compare an event log and a corresponding process model. In this paper, we propose an entirely new approach to conformance checking based on neural network-based embeddings. These embeddings are vector representations of every activity/task present in the model and log, obtained via act2vec, a Word2vec based model. Our novel conformance checking approach applies the Word Mover’s Distance to the activity embeddings of traces in order to measure fitness and precision. In addition, we investigate a more efficiently calculated lower bound of the former metric, i.e. the Iterative Constrained Transfers measure. An alternative method using trace2vec, a Doc2vec based model, to train and compare vector representations of the process instances themselves is also introduced. These methods are tested in different settings and compared to other conformance checking techniques, showing promising results.

Process mining aims to provide insights into the actual processes based on event data. These data are often recorded by information systems and are widely available. However, they often contain sensitive private information that should be analyzed responsibly. Therefore, privacy issues in process mining are recently receiving more attention. Privacy preservation techniques obviously need to modify the original data, yet, at the same time, they are supposed to preserve the data utility. Privacy-preserving transformations of the data may lead to incorrect or misleading analysis results. Hence, new infrastructures need to be designed for publishing the privacy-aware event data whose aim is to provide metadata regarding the privacy-related transformations on event data without revealing details of privacy preservation techniques or the protected information. In this paper, we provide formal definitions for the main anonymization operations, used by privacy models in process mining. These are used to create an infrastructure for recording the privacy metadata. We advocate the proposed privacy metadata in practice by designing a privacy extension for the XES standard and a general data structure for event data which are not in the form of standard event logs.

Predictive business process monitoring techniques aim at making predictions about the future state of the executions of a business process, as for instance the remaining execution time, the next activity that will be executed, or the final outcome with respect to a set of possible outcomes. However, in general, the accuracy of a predictive model is not optimal so that, in some cases, the predictions provided by the model are wrong. In addition, state-of-the-art techniques for predictive process monitoring do not give an explanation about what features induced the predictive model to provide wrong predictions, so that it is difficult to understand why the predictive model was mistaken. In this paper, we propose a novel approach to explain why a predictive model for outcome-oriented predictions provides wrong predictions, and eventually improve its accuracy. The approach leverages post-hoc explainers and different encodings for identifying the most common features that induce a predictor to make mistakes. By reducing the impact of those features, the accuracy of the predictive model is increased. The approach has been validated on both synthetic and real-life logs.

Predicting the next event(s) in Business Processes is becoming more important as more and more systems are getting automated. Predicting deviating behaviour early on in a process can ensure that possible errors are identified and corrected or that unwanted delays are avoided. We propose to use Bayesian Networks to capture dependencies between the attributes in a log to obtain a fine-grained prediction of the next activity. Elaborate comparisons show that our model performs at par with the state-of-the-art methods. Our model, however, has the additional benefit of explainability; due to its underlying Bayesian Network, it is capable of providing a comprehensible explanation of why a prediction is made. Furthermore, the runtimes of our learning algorithm are orders of magnitude lower than those state-of-the-art methods that are based on deep neural networks.

Nowadays predictive process mining is playing a fundamental role in the business scenario as it is emerging as an effective means to monitor the execution of any business running process. In particular, knowing in advance the next activity of a running process instance may foster an optimal management of resources and promptly trigger remedial operations to be carried out. The problem of next activity prediction has been already tackled in the literature by formulating several machine learning and process mining approaches. In particular, the successful milestones achieved in computer vision by deep artificial neural networks have recently inspired the application of such architectures in several fields. The original contribution of this work consists of paving the way for relating computer vision to process mining via deep neural networks. To this aim, the paper pioneers the use of an RGB encoding of process instances useful to train a 2-D Convolutional Neural Network based on Inception block. The empirical study proves the effectiveness of the proposed approach for next-activity prediction on different real-world event logs.

Predictive business process monitoring (PBPM) techniques predict future process behaviour based on historical event log data to improve operational business processes. Concerning the next activity prediction, recent PBPM techniques use state-of-the-art deep neural networks (DNNs) to learn predictive models for producing more accurate predictions in running process instances. Even though organisations measure process performance by key performance indicators (KPIs), the DNN’s learning procedure is not directly affected by them. Therefore, the resulting next most likely activity predictions can be less beneficial in practice. Prescriptive business process monitoring (PrBPM) approaches assess predictions regarding their impact on the process performance (typically measured by KPIs) to prevent undesired process activities by raising alarms or recommending actions. However, none of these approaches recommends actual process activities as actions that are optimised according to a given KPI. We present a PrBPM technique that transforms the next most likely activities into the next best actions regarding a given KPI. Thereby, our technique uses business process simulation to ensure the control-flow conformance of the recommended actions. Based on our evaluation with two real-life event logs, we show that our technique’s next best actions can outperform next activity predictions regarding the optimisation of a KPI and the distance from the actual process instances.

The present study investigates the relationship between IT organizational culture type and Business Process Management adoption in organizations implementing IT solutions. Specifically, the study investigates how the success of BPM adoption varies by organizational culture type. The target population consisted of IT resources who work in the United States at organizations with at least 50 employees and who have participated in the development and implementation of a BPM initiative involving an IT solution within the last two years. A survey was conducted with 157 anonymous participants representing the target population. The study found the highest level of BPM adoption success was with the adhocracy culture type compared to the market and hierarchy culture types. There is a significant positive correlation between the adhocracy culture type and BPM adoption as measured by BPO. There is also a significant negative correlation between the market culture type and BPM adoption as measured by BPO and PPI. The insights gained by this study can help practitioners make informed decisions on their BPM adoption approach within their IT community and scholars in future research on the relationship between organizational culture and BPM.

Process mining is proffered to bring substantial benefits to adopting organisations. Nevertheless, the uptake of process mining in organisations has not been as extensive as predicted. In-depth analysis of how organisations can successfully adopt process mining is seldom explored, yet much needed. We report our findings on an exploratory case study of the early stages of the adoption of process mining at a large pension fund in the Netherlands. Through inductive analysis of interview data, we identified that successful adoption of process mining requires overcoming tensions arising from discontinuing old practices while putting actions into place to promote continuity of new practices. Without targeted strategies implemented to transition users away from old practices, data quality is jeopardised, decision-making is impeded, and the adoption of process mining is ultimately hampered.

Digital innovation has forced companies to change some well-established business processes. Thus, incorporate information technology into business processes is not enough. We argue that both BPM role and related capabilities might need to be re-interpreted for the digital future. In this sense, the BPM discipline must identify relevant instruments for building on new ways to analyze, understand and support such transformations. This paper examines the results from a car rental company case in the light of DT&I-BPM-Onto, an ontology that encompasses relevant theory on the digital transformation domain. The company carried out a successful initiative by digitizing its primary end-to-end process that contributed to improvement of the company’s Net Promoter Score (NPS). The focus of the case study was the process’ redesign; thus we investigated: (i) the factors that led the decision to digitalize the process; (ii) how the digital transformation was conducted; (iii) the characteristics of the industry, the business itself, the transformed process, and the type of innovation implemented. The main contribution of this paper is to demonstrate that DT&I-BPM-Onto supports taking a broader picture of a process transformation case. Moreover, we provide insights and practical lessons for future projects and further research.

BPM maturity models (MMs) help organizations accomplish the BPM capabilities paramount for organizational success. Although much literature deals with how to design MMs, little knowledge exists of how organizations use BPM MMs. Moreover, the academic literature about MMs is scattered, making it hard for practitioners to learn from academia. Our purpose is to offer a holistic journey to guide organizations through three phases of BPM MM use, namely (1) choosing one out of many MMs that fits the organization’s context, (2) tailoring the MM to particular needs, and (3) advising during and after a maturity assessment. Starting from a synthesis of known guidelines, a framework for BPM MM adaption is presented with evidence of its applicability when organizations are conducting maturity assessments. The analysis calls for research to derive specific guidelines for different contexts, e.g., for different levels of maturity and/or when maturity assessments are driven by consultants.

Process standardization is for many companies a matter of strategic importance. Process standardization enables companies to provide consistent quality to customers and to realize returns of scale. Research in this area has investigated how process standardization impacts process outcomes, such as cycle time, quality, and costs. However, there are only limited insights into antecedents that lead to process standardization. Furthermore, it is not clear which contextual elements play a role when standardizing business processes. In this paper, we address this research gap by developing a theoretical model for business process standardization. The model is relevant for academics and practitioners alike, as it helps to explain and predict business process standardization by various antecedents and contextual factors.

Introducing change to organizational business processes is an inherently social event. People perform process activities to realize corporate and personal goals. When confronted with changes to their daily routine environments, people, being social actors, reflect critically on the changes presented to them. Collective interactions may lead to acceptance or rejection decisions about the process change, which could be why process change projects regularly fail during their implementation phase. Managing the complexity of interacting social mechanisms during the process implementation phase may be decisive in determining the success or failure of process change projects.

The lack of social mechanism models in this field indicates a business domain less managed. Problems during the process implementation phase may have financial implications, and can cause delays that reduce customer satisfaction.

This paper presents a research approach, consisting of a conceptual ontology model together with a mechanism discovery method. The approach seeks to uncover causal social mechanisms underlying adoption and rejection phenomena during business process implementation. It aims to strengthen research seeking to explain ‘why things happen’ during change initiatives.

The impact of this research envisions a central mechanism repository to further advance BPM practices. The mechanism repository together with the change ontology model could assist with the analysis of social dynamics from the people perspective to improve the management of process implementation projects.

In this study, we analyze the activity logs of fully resolved incident management tickets in the Volvo IT department to understand the handoff patterns i.e., how actors pass work from one to another using sequence analytics, an approach for studying activity patterns from event log sequences. In this process the process model itself is rather simple, but a large amount of variety is present in it in terms of the handoff patterns that arise. Hence, process modeling is not so helpful to gain a deeper understanding of the performance of the process. We offer an alternative approach to analyze such processes through the lens of organizational routines. A generic actor pattern here describes the sequence in which actors participate in the resolution of an incident. We characterize actor handoff patterns in terms of canonical sub-patterns like straight, sub- and full-loop, and ping-pong. Then, we predict incident resolution duration with machine learning methods to understand how actor patterns affect duration. Finally, the evolution of patterns over time is analyzed. Our results shed light on emergence of collaboration and have implications for resource allocation in organizations. They suggest that handoff patterns should be another factor to be considered while allocating work to actors along with position, role, experience, etc.