Business process simulation is a quantitative analysis technique of a business process. It enables organizations to analyze the performance of their process in an imitated virtual environment. Furthermore, it allows them to test different redesign scenarios safely to identify the effects of operational changes before implementation. In this talk, I will explain what business process simulation (BPS) is and the ingredients of a BPS model. We will talk about the creation of a BPS model via different techniques (manual, process mining, and deep learning) and its validation. Furthermore, we will look at the architecture of process simulation engines using Scylla as an example. In summary, we will walk through the simulation steps from model creation, validation, and execution to result interpretation.

The lecture takes place in A-2.1 from 1:30 pm to 3:00 pm.

Understanding the patient journey in a given disease is paramount to improving patient care. The state of the art in the industry is to conduct interviews or surveys with a few patients and physicians to generalize the patient journey. At mama health, Adriatik’s goal is to discover the patient journey from thousands of patient narratives by applying process mining. In this lecture, Adriatik focuses on the technical challenges and results of understanding the patient journey directly from patient experiences.

The lecture takes place in A-2.1 from 1:30 pm to 3:00 pm.

Just like every software component has an interface, every open software architecture has an application programming interface (API) which delimits its boundary, keeping it separate from other systems, while making it possible to connect and interact with them. Process-centric APIs are designed so that processes can flow through them. They are the main entry points for delegating execution, exchanging data, and transferring control back and forth across composable software systems. In this talk we illustrate a small collection of API design patterns derived from assigning the responsibility for determining “who does what when” to different parties, gradually deconstructing what it means to execute a process in a decentralized way.

Camunda 8 features a distributed event-driven process engine with eventual consistency and at least once semantics. Distributed applications have many advantages, e.g., polyglotism, horizontal scaling, and fault tolerance. But they need to deal with communication overhead, the absence of a central database, and the risk of potential inconsistencies. To use a distributed process engine to its fullest, applications and business processes may need to be redesigned: Services-tasks must be idempotent, and the transaction handling needs to be an explicit part of the process model. This lecture talks about the steps necessary to move to a distributed process engine as well as general Dos and Don’ts.

In the first lunch talk of the summer semester, Maximilian talked about his experience from supervising a team elicitating the travel reimbursement process at the HPI in the course of the Trends and Concepts in the Software Industry seminar with Professor Plattner. The goal for the team was to capture the as-is process including travel request and reimbursement, compare it to the best practice process provided by SAP Concur, and finally come up with aspects of the current process that constitute requirements for an adapted best practice process.

Anjo and Maximilian, two new PhD students at the BPT chair, participated in this year’s edition of the Central European Workshop on Services and their Composition (ZEUS) where they both submitted a position paper to outline their future research plans. In their talk, they presented their experiences in Hanover and the other papers presented at the workshop, concerning, for example, semantically equivalent elements in the BPMN standard, and BPMN layout analysis.

Reference management tools help researchers in collecting, organizing, and citing the literature they work through. In this session, Maximilian and Tom presented their tool of choice for managing the sources they use for their papers. Maximilian started with an introduction to Citavi, showing his usual workflow with the tool and its browser plug-in which makes the addition of new references as easy as the click of a button. Afterwards, Tom talked about his experience with Zotero, an open source tool available on most platforms. While both provide a similar set of features, getting to know their respective intricacies made this talk very interesting, especially for those not using a reference tool yet.

The plethora of process mining algorithms builds on directly-follows relations. Even though various improvements have been made in the last decade, there are serious weaknesses of these relationships. Once events associated with different objects that relate with a cardinality of 1:N and N:M to each other, these algorithms produce spurious relations, self-loops, and back-jumps. This is due to the fact that the event sequence as described in classical event logs differs from event causation. This talk addresses the research problem of representing the causal structure of process-related event data and presents an approach for causal process mining.

In the first session of the revived BPT Lunch Talks, Jonas talked about his impressions from the Hawaii International Conference on System Sciences, accompanied by a set of (physical) photos from his visit. The impressions ranged from inspiring research talks to a volcano trip and flightseeing with a helicopter.

The increasing volume of event data that is recorded by information systems during the execution of business processes creates manifold opportunities for process analytics. At the same time, increasing data volumes impose challenges regarding the scalability of analysis techniques, especially in the context of exploratory analysis. Observing that many insights about a process may be derived from a small share of the event data available in common application scenarios, we provide an overview of our work on sampling in process mining, covering algorithms for process discovery as well as conformance checking.