The Inaugural BPM Expert Forum took place on March 2, 2021, where renowned BPM thought leaders Prof. Wil van der Aalst and Prof. Michael Rosemann provided their viewpoints on “BPM Past, Present and Future”. Below please find the recording of the event and a Q&A. Thanks to Wil and Michael for taking the time to answer the questions that were communicated during the BPM Expert Forum and that could not be answered there due to time constraints.
What about digital twins?
[WvdA] The term digital twin is used in many different ways ranging from computer-aided design to model-based systems engineering. When limiting the notion to Digital Twin of an Organization (DTO, as defined by Gartner) then it is clear that process mining will play a key role. Using process discovery, one can learn realistic models and simulation can be used to answer what-if questions. At RWTH Aachen University, we developed the concept of a digital shadow which provides a concrete realization of this idea.
What would be a consequence to DX of BPM just described?
[MiRo] Assuming that DX means developer experience and that this question refers to the increased nexus between processes and data, it leads to the profile of developers how need to have process and data literacy. The default process analyst is expected to have an ability to engage with various stakeholders and needs strong people-related skills. A world, however, that becomes data-intensive will increasingly see the emergence of inductive approaches (e.g., process mining). Thus, a developer in this environment needs to be driven by evidence as opposed to confidence. We also can expect a more continuous approach to process design and change as we react to new data sets whereas traditional process development is more based on staged models.
RPA is relatively quick and low-cost to implement but rarely factors in the cost of increasing the environment complexity and building on already broken processes. How do you seen the improvement focus of Process Mining and the detail/reality of RPA working together?
[WvdA] As mentioned in my talk RPA and process mining complement each other. Process mining provides and end-to-end/top-down view and RPA aims to automate specific tasks. Process mining can be used to learn these tasks, but this is just one of many use cases.
From your point of view, what is the roadmap or foundations to reach process mining and RPA?
[WvdA] Process mining and RPA are very close to the data and actual operations. Starting point is to have a team that has data management and data science skills. It is quite pointless to discuss about imaginary data. In many process mining projects, 80% is spent on data extraction and often pre-existing data quality problems get visible. This needs to be addressed first. Therefore, one should not think in terms of business cases to solve problems that need to be solved anyway.
Why everyone confuses between BPM and BPMS?
[WvdA] This is correct. BPM does not require BPMS, but BPMS requires BPM. It seems that classical BPM systems will play a marginal role in current developments. Therefore, the BPM field should focus on the process-aspects rather than imposing new systems.
Do you think BPM provides today the right philosophy and framework for digital innovation?
[MiRo] There are two perspectives on this topic. First, the process of innovation. You might regard the process of digital innovation as just another process that requires BPM’s attention. If so, what would be, for example, an appropriate reference process model or what systems could support the timely execution of process innovation? Second, consider the digital innovation of processes. This includes the increasingly sophisticated automation of a process (e.g., smart contracts, augmented work, robotic resources) as much as capitalisation on entire new design options (e.g., process individualisation via 3D printing, event brokering). The latter is in terms of our understanding and the related academic and professional knowledge still in its infancy.
What makes a process well designed? Are these criteria built into BPM tools?
[MiRo] In my presentation I briefly referred to a hierarchy of process needs. Foundational are engineering needs, i.e. the process is executable and reflects the design intentions. In addition, a process has to comply to various further design rationales, e.g., low cost-to-execute. BPM tools tend to focus on engineering needs (e.g., syntactical correctness), but lack often support for design criteria such as ease-of-use by the customer, carbon footprint or advanced compliance assessments.
How do we differentiate between BPMS and WFA
[WvdA] The idea of a BPMS is that one can model a process and then generate a system from it. This was too naïve. Classical WFM systems were even more limiting since the focus was on automating simple processes that can also be programmed. However, the uptake of RPA shows that that there will be a revival of automation using new, more cost-effective concepts. Let’s not try to replace SAP and all of the other systems that together form the IT-ecosystem of an organization. Let’s augment these systems with process mining and RPA.
Hi Will - Great to hear you again, as always. Can you also talk a bit about how organizations are seeing BPM capability - where do they typically house this capability - Within EA / with COO team or something different?
[WvdA] The problem is that things are scattered. Larger organizations may have lean, BPM, EA, and process mining departments coexisting. There need to be integrated rather than be in competition. As always, real process improvements “hurt”. Therefore, there needs to have the support of higher management.
From your huge academic experience, what are the real reasons why universities are still too far removed from business life? Do business people find academics too theoretical?
[MiRo] Universities vary in the extent to which they are engaged with industry partners, and vice versa. The incentive schemas in universities (e.g., encourage publications) do not directly encourage industry engagement, and industry partners might struggle to understand how to approach and engage with a university. The timeframes of both parties (industry: fast delivery/utility, university: rigorous assessment) are also not always aligned. Finally, governance and partnership models (e.g., account management) as well as language used might be further road-blocks. However, universities increasingly recognise that access to innovative research questions, empirical evidence as well as the opportunity to create real-world impact are further motivators for industry engagement.
How should organizations combine the three types of processes management approaches: exploration, exploitation, hibernation?
[MiRo] These types will be of different relevance depending on the sector and the specific process. Process exploration is relevant when there is a richness of opportunities (e.g., potential to scale up the process to different types of products) whereas process exploitation reflects the continuous process innovation idea plus more recently the idea of process switching (e.g., COVID-inflicted switch from offline to online). Process hibernation is needed when a process needs to be ‘put on hold’. Some organisations (e.g., retail) had to deploy in recent times all three forms at the same time.
Thank you great Professors for your insights, I am interested with decision making (Human & Data concerns) in organizations, how can I use BPM Knowledge ? Starting point? Thank you
[MiRo] Processes vary in their knowledge and data dependency. As discussed in Wil’s presentation, we move towards augmented process executions, i.e. decisions are made by the most appropriate mix of humans and machines (data & algorithms). There is great potential for you to deploy BPM knowledge to decision making processes and decision process management is a promising path to improve the quality of decisions and to accelerate decision making processes (to real-time decisions in case of completely automated decision processes).
Thank you for such an inspiring talk. Industrial partners are aware of the importance of business process management. Nevertheless, it is difficult to bring academic innovation and tools to practice. Do you think that industrial partners need to become “BPM implementation partners”? How to motivate them to adopt such an active role in innovation?
[MiRo] Change occurs in three stages, awareness, acceptance and action. Thus, I suggest starting with awareness. What is BPM and what are the benefits of deploying it? Second, ‘industrial partners’ need to assess the extent to which they should capitalise on these benefits (acceptance stage). Partnering with a university that has a profile in BPM can help to stay in touch with latest BPM capabilities and cases plus securing BPM-aware, new talent.
Thank you very much for the great presentations! What do you think is the role of disruptive technologies like machine learning in the field of BPM? And what are the big question marks in your opinion? Like what are the questions academics should address to advance the field?
[WvdA] Advances in machine learning have been spectacular. At the same time, the applicability of machine learning in the narrow sense (i.e., neural networks) is limited. One needs to have a lot of (labeled) data. The challenges for BPM are: (1) what tasks can be automated and supported by these new technologies and (2) how to support the BPM-function itself. To support BPM itself process mining is more relevant than machine learning in the narrow sense. It is important that academics look for original questions. It does not make any sense to try and make incremental improvements in areas where the industry is leading. I sense that a lot of brainpower is wasted on generating more of the same. Whatever questions we look at, it has to be process-specific.
Do process architecture and role activity still hold relevance in the futuristic abductive BPM?
[MiRo] Yes, they do. In addition to the existing and relevant architecture and roles, abductive BPM will lead to desired architectures (e.g., an architecture for an ecosystem). The focus of the activity will be then to design a path towards this architecture. BPM professionals with strong abductive skills need to be familiar with essential design and experimentations skills as (unlike deductive and inductive BPM), abduction is an iterative process.
Currently these two are two sizes of the same coin but educated and exposed in totally separated worlds… Do you foresee a merger of BPM techniques with data modelling techniques? If the data structure is right, the better the process analysis and process mining…
[MiRo] No doubt, process science and data science, and with it related techniques and modelling approaches, will move even closer together as (1) processes are becoming more data-intensive and (2) data science becomes more process-aware. Deploying sound data structuring and analysis to process mining is essential, and the current focus of process mining which has matured substantially since the early days of often poorly structured event logs.
In general, I agree with both of you. But Wil, I think there is value in modeling for specific purposes. Do you agree, and can you characterize those? Case in point: why did SAP buy Signavio – for its process mining capabilities?
[WvdA] SAP always had problems embracing process management. SAP was happy with organizations like IDS Scheer and tools such as ARIS to do this. However, there was a huge gap between both. A nice illustration is the SAP reference model, which had little to do with the actual SAP system and its usage. In hindsight, traditional process modeling approaches and debates on process notations completely missed the point. I tried to explain this in my talk and many will not like it. During the last decade, SAP left the process management market to companies like Celonis and Signavio. With the uptake of process mining and other data science techniques, data-driven process management is no longer a small niche, but the main driver for innovation. Signavio was first exclusively focusing on modeling, but in recent years, they realized that this would not be viable in the long run. Therefore, they invested heavily in process mining, leveraging the process modeling capabilities to support conformance checking. The next challenge to integrate Signavio’s BPM capabilities in SAP. This will not be so easy. Moreover, SAP is increasingly used in a landscape where also many other systems are used. Therefore, data integration problems will continue to exist.
Could you please share advice for the very first process mining project to take in an organization. Thank You.
[WvdA] I made a slide set for organizations to do a self-assessment first, see here. The main problem is that people talk about data and processes in a very imprecise way. Process mining requires event data and process-related questions. It is a tool as generic as an Excel spreadsheet and can be used in any organization. However, when using the tool, one should be able to extract the right data and ask the right questions. This is far from trivial. Projects are doomed to fail when people have difficulties finding the data or ask questions unrelated to the data they have.
Prof.dr.ir. Wil van der Aalst is a full professor at RWTH Aachen University leading the Process and Data Science (PADS) group. He is also part-time affiliated with the Fraunhofer FIT where he leads FIT’s Process Mining group and a member of the Board of Governors of Tilburg University. Wil van der Aalst has published over 800 articles and books and is typically considered to be in the top-15 of most cited computer scientists with an H-index of over 155 and more than 110.000 citations. Next to serving on the editorial boards of over ten scientific journals, he is also playing an advisory role for several companies, including Fluxicon, Celonis, and UiPath. Van der Aalst is an IFIP Fellow, IEEE Fellow, ACM Fellow and received honorary degrees from the Moscow Higher School of Economics (Prof. h.c.), Tsinghua University, and Hasselt University (Dr. h.c.). He is also an elected member of the Royal Netherlands Academy of Arts and Sciences, the Royal Holland Society of Sciences and Humanities, the Academy of Europe, and the North Rhine-Westphalian Academy of Sciences, Humanities and the Arts. In 2018, he was awarded an Alexander-von-Humboldt Professorship.
Prof. van der Aalst will provide his insights on what did we miss in the last several decades of BPM research and development. He states that “Traditionally, BPM is about process modeling and BPM/WFM systems. In the 1990-ties we assumed that we could generate information systems by modeling processes properly. However, this turned out to be the exception rather than the rule. Instead, Process Mining (PM) and Robotic Process Automation (RPA) are booming while organizations are still using ERP systems from SAP, Oracle, Microsoft, Infor and the like. Therefore, we need to rethink BPM and no longer assume we can improve processes by just modeling them. We need to exploit the data available, go beyond simplistic BPMN models, and accept that BPM needs to build on existing complex data-centric applications. This will help us to expand the scope of process management far beyond contemporary BPM niches.”
Prof. Dr. Michael Rosemann is a full professor at Queensland University of Technology leading the Centre for Future Enterprise. He has published comprehensively on various aspects of BPM and initiated research in areas such as BPM maturity models, context-aware BPM, ambidextrous BPM and trust-aware BPM. His books are available in five languages and he serves on the editorial boards of 10 international journals. Rosemann has conducted invited keynotes at academic and professional BPM conferences and is a regular advisor to board rooms across various industries. He is also the Honorary Consul for the Federal Republic of Germany in Queensland, Australia.
Looking forward, Prof. Rosemann highlights the need to understand what we might miss in future BPM related research and developments. He has indicated that “The management of business processes has been mostly centred on the identification of process issues and been guided by process improvement approaches such as lean management, six sigma or process automation. However, such reductionist approaches to process management are no longer sufficient. An opportunity-rich environment and the changing nature of processes (e.g., transformational, scalable, entrepreneurial, socially responsible) require entire new BPM approaches. Deductive and also inductive (e.g., process mining) methods need to be complemented by abductive approaches enabling advanced process design. This will require new design criteria, methods, systems and most of all mindsets for BPM to stay relevant.”