PI(s): Dr. Sarma Vrudhula
Problem description, motivation, and objectives: Edge devices increasingly need to execute advanced AI workloads locally to reduce latency, and energy consumption. However, emerging models such as Kolmogorov–Arnold Networks (KANs) introduce
from Transprecision Computing for Improving Energy Efficiency of ML Networks
PI(s): Nakul Gopalan
A significant challenge in problem domains like agriculture and construction is that the environment is unstructured and presents significant variety of hazards on site. Such lack of structure and hazard
PI(s): Sarma Vrudhula (ASU), Mehdi Ghasemirahaghi (SIUC)
Problem description, motivation, and objectives: The goal of this project is to enable the efficient execution of AI workloads on a set of distributed edge devices pursuing the same mission.
from RL-Based Decision Making of Distributed Systems under Degraded and Limited Network Conditions
PI(s): Dr. Giulia Pedrielli
Problem description, motivation, and objectives: Testing and validation of autonomous systems—particularly perception and decision modules in autonomous vehicles—require the ability to generate a large number of realistic yet safety-critical scenarios.
PI(s): Aman Arora
Defense systems must integrate legacy electronics (often running at 3.3 V–5 V and older packages) with modern subsystems (≈1.2 V–2 V). Lack of flexible, trusted, U.S.-fabricated devices that can programmatically bridge these voltage domains drives cost,
from Voltage Translation FPGA Design and EDA Flow for Countering Obsolescence
PI(s): Arindam Sanyal
Effective power management is essential for balancing energy efficiency and performance in modern microprocessors. Fully integrated digital low-dropout (DLDO) regulators provide a scalable and compact solution; however, conventional architectures such
from Reinforcement Learning-Based Digital LDO for Fast and Adaptive Transient Response
PI(s): Sarma Vrudhula
In resource-constrained edge devices, executing applications locally is increasingly necessary, as transmitting data to centralized servers adds latency, energy overheads, and security risks. To mitigate these issues, edge platforms are
from Circuits and Architectures for Low-Power Transprecision Computing on Edge Devices
PI(s): Spyros Tragoudas
Augmentations in a deep neural network (DNN) are introduced to enhance image classification accuracy while significantly reducing inference time. The architecture comprises a hybrid framework that combines binary classifiers based
from DNN Architectural Enhancements with Inference Time and Energy Considerations
PI(s): Nakul Gopalan
Farmers possess significant knowledge about crops, their distress, and diseases. However, it is labor intensive for farmers to notice all the plants at all times. There needs to be an
PI(s): Haibo Wang, Jun Qin, Spyros Tragoudas
As an important application in precision agriculture, variable-rate spray helps increase yield, reduce cost, and minimize environmental contamination. Canopy volume measurement units are critical sensing components in variable-rate spray systems.
PI(s): Hokeun Kim
This project aims to explore the design space of a cost-effective, versatile, and modular hardware/software (HW/SW) platform for supporting a diversity of requirements in agriculture technology (AgTech) through monitoring, sensing, controlling machinery, and communication.
PI(s): Arindam Sanyal
The DoE Office of Nuclear Physics seeks new developments in detector electronics with significantly improved energy, timing resolution, sensitivity, dynamic range (DR), and signal-to-noise ratio (SNR).
from Radiation Hardened Intelligent Readout ASIC for Nuclear Physics Experiments
NSF INDUSTRY-UNIVERSITY COOPERATIVE RESEARCH CENTER