on July 1, 2025 at 2:00 pm
01 July 2025 at 2.00 pm
Short Bio Renato Mancuso is an associate professor in the Computer Science Department at Boston University and the director of the BU Cyber-Physical Systems Lab. Renato received his Ph.D. from the Department of Computer Science at the University of Illinois at Urbana-Champaign in 2017. Since 2013, he has published over 80 papers in major, conferences, workshops, and journals. His work has received multiple awards, including multiple Best Paper, Oustanding Paper, Best Demo, and Best Presentation awards. Prof. Mancuso also received the NSF CAREER Award and the Gerald and Deanne Gitner Family Award for Innovation in Teaching with Technology in 2023, and the ACM SIGBED Early Career Researcher Award in 2024. His research has been supported by federal agencies and industrial partners such as Bosch, Red Hat, Cisco, and Xilinx. He is a senior member of the IEEE and the ACM SIGBED.
Abstract We entrust our lives to computing systems. Whenever we fight a disease in an intensive care unit, brake to avoid a car collision, or hop on a flight, our lives depend on a computing system doing the right thing (logical correctness) at the right time (timeliness). Timeliness is key for safety-critical systems as they interact with the physical world. As our modern lifestyle becomes symbiotic with computer systems, we also expect timeliness from the graphical interfaces in our mobile devices, data streaming services, and edge cloud systems.
Unfortunately, timeliness is a property that cannot be explicitly programmed, as it emerges from the interplay of software and hardware components. Without explicit control over one such interplay, timeliness remains an elusive dimension. Adhering to the principle that “you cannot control what you cannot observe”, can we move away from workload-agnostic systems? Can we leverage fine-grained knowledge of said interplay to regain control over timeliness in our systems?
In this talk, I will outline the vision for “knowledge-driven system management”. Said vision is built upon a symphony of techniques to collect, retain, and leverage information about the low-level interactions between applications and hardware components. First, I will review just how much the lack of control over performance-critical system resources poses a significant barrier to complex yet safe systems. Next, I will introduce cutting-edge enabling mechanisms to gain progressively more insights and control over data flows between hardware components in embedded systems-on-a-chip. Finally, I will showcase how live tracking of applications progress can be exported and leveraged to construct proactive system management strategies.
Amphi Lebon - Polytech'Lille
français
English