Engineering Research and Innovation Seed Funding Program
To encourage new research collaborations, the College of Engineering Associate Dean for Research has initiated the Engineering Research and Innovation Seed Funding program (ERISF). The overarching goal of the ERISF program is to assist engineering researchers in developing new innovative research programs that have strong potential for significant future support from government agencies, corporations, industry, consortia, or foundations. Proposals must develop research initiatives with high potential for significant extramural funding.
CY2020 ERISF Projects
Principal Investigator: Xiangbo Meng, Mechanical Engineering
Collaborators: David Huitink, Mechanical Engineering
Emerging applications including electric vehicles, wind generators, solar converters, aerospace power conditioning, etc. are driving an ever-increasing need for high-temperature dielectric materials. In these systems, power systems and electronic devices have to operate at elevated temperatures (>140 oC), where current dielectric materials cannot survive. In this project, we propose to rationally develop novel organic-inorganic hybrid dielectric materials with desirable dielectric properties. To this end, atomic and molecular layer deposition (A/MLD) will be combined for developing new high-temperature dielectric materials. This high-risk investigation holds great potential to revolutionize power electronics through improved capacitor design. To develop and evaluate the proposed hybrid materials, Dr. Xiangbo Meng and Dr. David Huitink from Department of Mechanical Engineering will form a highly complementary team. This new collaboration is intended to deliver a new research capability and it is also expected that the synergic activities between the two groups will make University of Arkansas a leader in Energy, Materials, and Electronics.
Principal Investigator: Sarah Nurre, Industrial Engineering
Collaborators: Kelly Sullivan, Industrial Engineering and Benjamin Runkle, Biological Engineering
Food and agriculture (ag) is a critical infrastructure making up almost 20% of the economic Activity in the US. In spite of its importance, researchers have overlooked food and ag when Classifying infrastructure interdependencies. We propose to fill this need by identifying and Classifying the interdependencies of the food and ag sector within the state of Arkansas (AR) with the aim of strengthening its resilience. Ag is the top industry in AR wherein rice and poultry comprise two of the top commodities. Using rice and poultry as case studies, we will interview key personnel in the food and ag industry such as farmers, packaging managers, researchers, and members of the AR Division of Ag. Complementing this effort, we will interview managers of interdependent infrastructures (e.g., water). We will augment the interview data with data from GIS and literature to classify and create visualizations for the types and characteristics of interdependencies within food and ag in AR.
CY2021 ERISF Projects
Principal Investigator: Han Hu, Mechanical Engineering
Collaborators: Justin Zhan, Computer Science and Computer Engineering
Boiling is central to the high-performance cooling of nuclear reactors and high-power electronics, whereas the boiling crisis with a sudden and rapid reduction in heat transfer coefficient has raised significant safety concerns. The proposed project aims to enable early detection and prevention of the boiling crisis by leveraging the convolutional long short-term memory-based sequential classification. We hypothesize that sequential classification will lead to much higher accuracy and faster response than traditional image/thermal data analysis by capturing the dynamics in the image sequences. Furthermore, by extracting the key features of the boiling image sequences and correlating them with the thermal performance data, we aspire to obtain new physical insights into the mechanism triggering the boiling crisis. This collaborative effort will integrate the two PIs’ expertise in thermal engineering and data science to solve long-standing thermal challenges and promote the applications of advanced data analytics in engineering applications.
FY2023 ERISF Projects
Principal Investigator: Young Hye Song, Biomedical Engineering
Co-PI: Jin Woo Kim, Biological & Agricultural Engineering
This project seeks to develop and characterize cellulose nanocrystal-enhanced collagen scaffolds for peripheral nerve repair. Despite advances in biomaterials science and engineering, peripheral nerve injury remains a significant clinical challenge. Current clinical products have limitations, especially lack of nerve-mimetic physicochemical cues that promote axonal regeneration. As such, we will create biocompatible nerve grafts that provide topographical guidance to nerve regeneration. Aligned collagen fibers will mimic nerve matrix architecture, while cellulose nanocrystals will provide bioactive cues and modulate mechanical properties of the grafts. Pro-regenerative adipose-derived stem cells will be incorporated as a living supply of neuro-regenerative factors to improve nerve repair and regeneration.
Principal Investigator: Shengfan Zhang, Industrial Engineering
Co-PI: Han Hu, Mechanical Engineering
The project aims to develop and characterize a data-driven control algorithm for cooling systems to mitigate overheating-induced failures and reduce energy consumption. Reinforcement learning will be integrated with multimodal sensing to enable automatic active control of thermal management of electronic and power systems. Compared to the conventional proportional–integral–derivative controllers, the proposed algorithm is promising in handling high-frequency processes, e.g. thermal regulation under pulsed heat loads, mitigation of interpulse instability in GaN power amplifiers in radar systems. The proposed framework can be readily implemented in existing thermal management systems by retrofitting existing systems with an extra AI control unit.
Principal Investigator: Ben Runkle, Biological Engineering
Co-PI: Thi Hoang Ngan Le, Computer Science & Computer Engineering
Precision agriculture is a key strategy to enhance agricultural output and sustainable food production, and it relies on accurate estimates of spatial heterogeneity in the outdoor landscape. While humans (by foot or tractor) or flying drones can assess some heterogeneity, methods that support assessment within the crop canopy still need development. Advanced automation robotics can help assess this part of the landscape, but robots must be designed with care to balance rigor in the outdoor landscape with economic and accuracy considerations. To take advantage of recent robotics developments and resolve this challenge, Runkle and Le are combining diverse expertise to blend real-world use cases with cutting edge robot and artificial intelligence techniques. They will work with graduate students from each lab to design, test, and prototype robot solutions to form the basis of several federal grant opportunities and manuscripts, as well as train students in cross-disciplinary methods.
Principal Investigator: Dongyi Wang, Biological & Agricultural Engineering
Co-PI: Wan Shou, Mechanical Engineering
The poultry industry is the No.1 meat industry worldwide and the U.S is the world's primary poultry producer. However, many processing steps in the poultry industry still heavily rely on manual labor. Changing this costly workflow for resilient food productions becomes more urgent now than ever with many food companies continuing to struggle with the COVID pandemic. The challenges in automating these processes fall into intelligently handling soft meat products with various visual and mechanical characteristics. The team proposes a novel multimodal-sensor (visual+tactile) guided robotic chicken handling system that meets the requirements for integrated, autonomous, and resilient chicken processing. The specific seed fund objectives include: 1) Customize a low-cost high-density tactile sensor array embedded robotic gripper. 2) Develop a high-resolution and high-speed 3D imaging system through an offline-online co-design strategy. The outputs from this study will offer reliable tactile and visual information for further realizing adaptive and integrated robotic control.
Principal Investigator: Yanjun Pan, Computer Science & Computer Engineering
Co-PI: Jingxian Wu, Electrical Engineering
Anonymity has been recognized as an essential attribute in privacy-preserving communications for a wide range of applications, such as smart grids, Internet-of-Things (IoT), autonomous driving, etc. Stimulated by 5G networks and beyond, a plethora of new services and infrastructures are expected to bring billions of new wireless devices to the network, which imposes unprecedented challenges on the security and privacy of communication systems. Nevertheless, existing anonymous communications systems only address authentication and routing at upper layers, ignoring the fact that the physical (PHY) layer also contains a privacy preserving link. In particular, the radio frequency (RF) fingerprints that capture unique, humanlike device discrimination have been envisioned as a promising solution for autonomous device authentication. The security of RF fingerprints-based approaches stems from the common belief that reproducing or replaying a device’s RF fingerprint is practically hard. Yet, the correctness of this concept is still under-examined and can lead to potentially significant privacy leakage at the PHY layer. This proposal aims to conduct a comprehensive investigation covering various aspects from examining RF fingerprinting based security schemes to enhance cross-layer network anonymity.
It consists of three research thrusts:
- Revisiting RF fingerprints reproducibility through deep learning technologies
- Designing cross-layer anonymous communications systems
- Extensive experimental validation on proposed algorithms and protocols with a newly developed 5G communication testbed