Computer Science Department Colloquium
Human-Centered Intelligent Sensing Systems
Monday, February 22, 2021, 02:30pm - 04:00pm
Speaker: Jorge Ortiz
Jorge Ortiz is an Assistant Professor at Rutgers University where he directs the Cyber-Physical Intelligence Lab (CyPhy-Lab) and is where he is also a member of the Wireless Information Network Laboratory (WINLAB). His work focuses on building and studying sensing systems that learn about human behavior, from human feedback, and with humans to improve system objectives and enhance people’s lives. These include a broad range of sensing systems including smart objects, smart built environments, and smart cities, more broadly. Prior to joining Rutgers in 2018, he was a Research Staff Member at IBM Research working on machine learning and the internet of things. In the five years he was at IBM, he attained 12 patents and published in multiple top academic conferences, journals, and books and was awarded ‘Best Poster’ at IEEE/ACM IPSN '08, two ‘Best Paper Runner-ups’ at Buildsys '15, ‘Best Paper’ at ICISSP '18, and ‘Best Paper Runner-up’
at IoTDI '19. At IBM he led teams to commercialize two major research projects and bring them to market. Dr. Ortiz also has extensive industry experience, which includes several years at Oracle Corporation and has worked at and led multiple startups. He is currently serving as co-TPC chair for Buildsys 2020. Dr. Ortiz attained his Ph.D. in Computer Science from UC Berkeley in 2013, M.S.
in Computer Science from Berkeley in 2010, and a BS in Computer Science from MIT in 2003.
Location : Via Zoom
Event Type: Computer Science Department Colloquium
Abstract: This talk is about the design of systems and algorithms for sensing systems that interact directly with humans. I will discuss our work in the Cyber Physical Intelligence lab (CyPhy-Lab) at Rutgers University where we study sensing systems that learn about human behavior, from human feedback, and with humans to improve system objectives and enhance people’s lives. I will describe three on-going projects that explore these themes more precisely. First I will describe PillSense, a smart pillbox system for medication adherence. This system helps us learn about how humans take their medication and how physical design is tightly coupled to the system’s ability to identify users effectively. We describe two pill box designs and associated algorithms to address the challenges posed. Then, I will discuss our on-going project Maestro, a system that learns from humans to fill the semantic gap between sensor measurements and their interpretation, in order to facilitate the construction of smart ambient-sensing applications in buildings. Finally, I will discuss project Paz, a system that attempts to learn when to interact with humans as we work to facilitate agent-human collaboration to both attain system objectives (i.e. efficiency) and enhance human productivity, comfort, and entertainment.
Contact Host: Dr. Richard Martin