Gen AI enhances BPM but cost, model complexity hinder AI adoption – PEX Research & Reports News
Community care neglected in digital transformation efforts, new opportunities to study processes through multiple data sources and the application of process trees
Add bookmark
PEX Network’s weekly news bulletin rounds up the latest research, reports and publications in operational excellence (OPEX), digital transformation, artificial intelligence (AI) and automation, business process management (BPM), process mining and process intelligence and more.
This week includes:
- Generative AI a significant facilitator of BPM
- Cost and model complexity barriers to enterprise AI
- Researchers examine the application of process trees
- Digital transformation efforts neglect community care
- New opportunities to study processes through multiple data sources
Don't miss any news, updates or insider tips from PEX Network by getting them delivered to your inbox. Sign up to our newsletter and join our community of experts.
Generative AI a significant facilitator of BPM
A new research paper explored the role of large process models (LPM) as a central conceptual framework for software-supported BPM in the era of generative AI. LPMs combine the correlative power of generative AI large language models (LLMs) with the analytical precision and reliability of knowledge-based systems and automated reasoning approaches, according to the researchers.
The chief aim of the research was to provide a balanced, feasibility-oriented discussion of the expected impact of foundation models on BPM software. “In this vision, the proposed LPM would enable organizations to receive context-specific (tailored) process and other business models, analytical deep-dives, and improvement recommendations,” the authors wrote. “As such, it would allow to substantially decrease the time and effort required for business transformation, while also allowing for deeper, more impactful, and more actionable insights than previously possible.”
The research concluded that implementing an LPM is feasible and has the potential to substantially facilitate BPM. “We see the LPM as a fusion of generative AI and traditional symbolic and statistical approaches to automating reasoning and decision-making in BPM,” the researchers stated. “Here, we expect a substantial industry impact over the next years.” However, they also highlighted limitations and challenges that need to be solved to implement particular aspects of the LPM vision.
Cost and model complexity barriers to enterprise AI
Cost and model complexity remain barriers to enterprise AI adoption,according to a new report from the IBM Institute for Business Value. The report, based on a survey of US-based executives, aimed to provide CEOs with actionable insights to make informed decisions about AI implementation and optimization. Of those surveyed, 63 percent said cost is the primary obstacle to generative AI adoption, while 58 percent cited model complexity as a top concern.
“Why is cost such an important consideration? Because it can vary widely depending on the model being used,” IBM stated. For example, larger models come with more data storage and compute costs, which can result in higher cloud-related bills. Large models also require more frequent updates, fine-tuning and maintenance, which come with talent costs.
Meanwhile, as both data volumes and model complexity increase, tech infrastructure must be able to handle the heavier load, IBM added. “Then there’s the matter of scaling. As more teams use generative AI in all its forms, organizations need to evolve their infrastructure or cloud environment to meet increased demand.”
Researchers examine the application of process trees
A new academic paper examined the development of novel, time-efficient decomposition strategies for the application of process trees – a type of process model often used in process mining. “Despite their frequent use, little research addresses the formal properties of process trees and the corresponding potential to improve the efficiency of solving common computational problems,” the authors wrote. Image left: Example of a process tree model. Source here.
The paper therefore sought to propose an invertible state space definition for process trees and demonstrate that the corresponding state space graph is isomorphic to the state space graph of the tree’s inverse.
“Our experiments confirm that our state space definition allows for the adoption of bidirectional state space search, which significantly improves the overall performance of state space searches,” according to the researchers.
Digital transformation efforts neglect community care
Better digitally enabled care in the community has the potential to improve quality of life for many people who have ongoing care needs including living with neurodiversity, physical disabilities and complex long-term conditions. However, digitally enabled community care in the UK is currently limited by low levels of investment and siloed approaches, according to a new report from The King’s Fund. “There are examples of good implementation and use of technologies, but these need to become the norm not the exception,” the report stated. To capitalize on future technologies, there needs to be “substantial improvements” to digital infrastructure and the use of basic technologies, in addition to growing the capability and capacity in community settings so that technology and new skills can be integrated, it added.
“More work needs to be done on this topic to drive forward digitally enabled care in the community.” The King’s Fund recommended work in the areas of a national vision to guide local decisions, strategies for co-developing digitally enabled services, shifting culture to embrace the public’s digital capability, supporting staff to embed technology in their roles and creating cross-NHS and social care environments that enable innovation.
New opportunities to study processes through multiple data sources
A new paper introduced and conceptualized a new scientific field: process science. Process science is concerned with the capturing and understanding of socio-technical processes of different kinds aiming to inform interventions to and the design of processes, especially by leveraging digital trace data and computational techniques, the authors stated. “The ubiquitous availability of digital trace data, combined with advanced data analytics capabilities, offer new and unprecedented opportunities to study such processes through multiple data sources,” they wrote.
The next important step is to start bringing process science to life and begin research projects that embrace and advance the field, the authors concluded. “Process science is in the making. Everyone who wants to engage with it is welcome to shape the field as it evolves.”
Strategic Transformation & Operational Excellence Europe 2024
Meet and learn from the leaders and experts in the field of operational excellence, business process and digital transformation. Discover the latest tools and best practices for achieving customer-centricity, agility, and innovation. Register now and join the most influential community of business transformation and process professionals in Europe.
To view this content, please fill out the form to register and become a member.
Or, if you're already a member, sign in below to view.
Please note: That all fields marked with an asterisk (*) are required.
