The Bridge - Issue 2, 2022 - 33

Graduate Research SP
Katelyn Brinker
Gamma Theta, Iowa State University, Ph.D. Candidate in Electrical Engineering
RESEARCH TOPIC
Chipless RFID for Structural Health Monitoring (SHM)
Radio Frequency Identification (RFID) systems typically consist of a reader and a tag. The tag can provide identification information
and/or information about the environment, such as temperature and humidity. Chipless RFID is a subset of the RFID field where
these tags don't have any power source or electronics. Instead, their information is " stored " in their structure - that is they're designed
to scatter in a specific way when interrogated with an electromagnetic wave. This scattered response can be viewed in the time-,
frequency-, or spatial-domain and a binary code can be assigned to it. Changing the structure of the tag or the environment that
the tag is in, produces changes in this binary code, and in this way, identification or sensing can be performed. The wireless passive
nature of this technology results in less expensive and less invasive sensors, which can allow for more comprehensive monitoring
of structures.
Katelyn's work, supported by a NASA Space Technology Research
Fellowship, has more specifically examined how chipless RFID
tags can be used for structural health monitoring applications in
the space and aerospace realms. She has developed new tags
that can be used for microwave materials characterization and
rotation sensing, examined inkjet-printing for tag manufacturing,
and developed new measurement methods to address some of
the current challenges in the chipless RFID field.
Chipless RFID tags and custom 3D printed antenna for tag measurement.
LEARN MORE
https://scholar.google.com/citations?user=scaP5rQAAAAJ&hl=en
Shail Dave
Epsilon Beta, Arizona State University, Ph.D. Student in Computer Engineering
RESEARCH TOPIC
Design Space Modeling and Optimizations for Dataflow Accelerators
Achieving efficient processing of commonly used computing applications is important, e.g., for improved performance and energy efficiency.
Therefore, domain-specific hardware accelerators have been increasingly designed, especially for tasks involving intelligence through
machine learning (ML). Techniques are required for mapping the application tasks on these hardware accelerators in a resource-efficient
manner. Additionally, tools for modeling the accelerator's execution and its design exploration are also needed, and help the designers in
characterizing the execution efficiency and improving it further.
Shail's research include sustainable techniques for execution modeling, task mapping, and agile design space exploration for reconfigurable
dataflow accelerators. For instance, he has developed automation techniques for obtaining efficient mappings of ML operators and
performance-critical loops of general-purpose applications on dataflow accelerators. Further, ML models can be sparse and compact, such
that high task accuracy can be obtained at
lower computing, communication, and memory
requirements. Exploiting such benefits, however,
usually requires additional hardware/software
support. Shail's recent work, including a literature
survey published in Proceedings of the IEEE,
analyzes different techniques for accelerating
such sparse and compact ML models and their
execution efficiency.
1- Efficiently processing ML models on dataflow accelerators requires succinct modeling and optimized
exploration of hardware/software; 2-Agile methodology for designing efficient accelerators with automated
full-stack development.
LEARN MORE
https://scholar.google.com/citations?user=SVft2R0AAAAJ&hl=en
CONTACT
https://sites.google.com/view/shail/
CONTACT
https://www.linkedin.com/in/katelyn-brinker/
TLIGHT
HKN.ORG
33

https://scholar.google.com/citations?user=scaP5rQAAAAJ&hl=en https://www.linkedin.com/in/katelyn-brinker/ https://sites.google.com/view/shail/ https://scholar.google.com/citations?user=SVft2R0AAAAJ&hl=en https://hkn.ieee.org/

The Bridge - Issue 2, 2022

Table of Contents for the Digital Edition of The Bridge - Issue 2, 2022