Search Results

This report documents the program and the outcomes of Dagstuhl Seminar 24292 "Improving Trust between Humans and Software Robots in Robotic Process Automation". The seminar dealt with topics targeted at developing frameworks and guidelines to empower the trust relationship between humans and Software Robots (SW) in Robotic Process Automation (RPA). RPA is a maturing technology that sits between the fields of Business Process Management (BPM) and Artificial Intelligence (AI). RPA allows organizations to automate high-volume and repetitive tasks – also referred to as routines – performed by human users. The enactment of these routines is emulated by means of a software (SW) robot that works on the applications' user interfaces (UIs) in the same way as the original human operators did. Recent research studies conducted on the effectiveness of RPA within organizations have found that implementation of SW robots does not always lead to the assumed effect, and many SW robots are subsequently withdrawn. In consequence, the human workforce takes over robotized tasks to perform them manually again and, in practice, replaces back SW robots. The fact is that integrating RPA into a human workforce alters the role of human employees and dynamics within the workforce, fueling a lack of trust in RPA technology, an issue deemed increasingly significant given its widespread use in many working domains. In this direction, this Dagstuhl Seminar aimed to bring together leading experts from industry and academia engaged in diverse communities related to RPA, including BPM and Human-centered AI, intending to reflect on the current RPA principles, which fail to deliver sufficient attention to the interplay between the human workforce and SW robots. The overall goal was to explore the scientific and technological foundations to pioneer new trust-aware RPA solutions that work in partnership with the human workforce, to enhance human capabilities rather than replace human intelligence and break through the barriers to human trust using RPA. The seminar outcomes will serve as a basis to foster joint research efforts and collaborations for charting a roadmap for future RPA research.

Business Process Management (BPM) has significantly advanced and gained high popularity in industry. However, it remains an open issue why tools frequently are used for business process modeling that are not mainly implemented for this purpose. Often, macros for Microsoft Visio or Microsoft Excel form the first choice to capture the flow of business activities. One reason why these tools might be used is the low training effort and the fast creation of a quick model, which can be generated with these tools. Another reason for the “lower” preference of BPM software tools might be their inability to respond to changes in technology and working styles, e.g. the shift towards "agile" processes and the "flattening" of workforce hierarchies that bring more stakeholders into contact with a much broader array of processing steps than before. A central question is whether the BPM community should create an entirely new paradigm for process modeling. One can think of more intuitive drawing conventions that laymen would use, and of models of an entirely different kind (i.e. not process-centric and not data- or case-centric) that still bear the possibility to support modern and future business process. The purpose of this seminar was to bring together a cross-disciplinary group of academic and industrial researchers to foster a better understanding of how to ease the access to, and applicability of, business process modeling. We discussed business process modeling approaches against emerging trends such as Internet of Things, the need for incremental and agile creation of new processes, and the need for workers to understand and participate in multiple contextual levels (e.g. transactional, business goals, strategic directions) while performing processes. The seminar also considered how new technologies, such as modern tools for UI design (e.g. D3, node.js) could be applied to support fundamentally shifts in how processes are modeled and how humans are involved with their execution.