Ezra's Round Table / Systems Seminar: Joseph Pekny (Purdue) - The Role of Knowledge Acquisition Frameworks and Deliberate Innovation in Addressing Intractability and Rapid Cycle Innovation

Description

The Process Systems Engineering (PSE) community has evolved in parallel with the action of Moore's Law for about 50 years. As computing has become millions of times more powerful/cost effective and changed qualitatively to include the internet, smart phones, virtual reality, the-internet-of-things, and self-driving cars the importance and application of PSE involving modeling, algorithm engineering, analytics, optimization, and process engineering has grown tremendously. Despite significant progress, some fundamental challenges have been recurring and resistant to a variety of approaches. In fact, the PSE framework is complementary to the rapid advances in artificial intelligence and pattern recognition, especially for the conceptual and engineering tools that can be brought to bear in autonomous machines and what they are capable of achieving. This presentation discusses the challenge of computational intractability in PSE in general and discusses the specific case of process planning and scheduling as an example that has defied systematic success and stymies autonomous manufacturing, but for which there is enormous value to intellectual advances. The concept of a knowledge acquisition framework is emerging as a useful means to address computational intractability and is supported by the next evolution in Moore's Law that will dramatically increase the amount of available fast memory economically available. Knowledge acquisition frameworks provide a means to conduct large scale practical experiments, provide value to users, and quickly apply learning to goal formation and their accomplishment. The knowledge acquisition framework underlying the Virtually Exhaustive Combinatorial System (VirtECS) System will be summarized as the product of more than 25 years of rapid cycle, practically driven research and development. This effort has historical parallels with the Deliberate Innovation approaches of the 20th Century and points to the critical role of prototyping and industrial partners to cutting-edge research.

Bio:
Joseph F. Pekny is a professor of chemical engineering at Purdue University. Professor Pekny was a founding director of Discovery Park and the Regenstrief Center for Healthcare Engineering. He has had numerous leadership roles at Purdue involving entrepreneurship, innovation, and startup of new activities. In addition, he cofounded and is currently CEO of Advanced Process Combinatorics, Inc. (APC)—a Purdue Research Park company. His research interests are in algorithm engineering for combinatorial optimization applications and in devising effective systems for translating research into practical impact. His work includes applications in energy systems modeling, research pipeline management, manufacturing, and planning and scheduling. He received his undergraduate degree in chemical engineering from Princeton University and his Ph.D. in chemical engineering from Carnegie Mellon University.