UMD Alum Wins ACM SIGMOBILE Dissertation Award

Nakul Garg speaking into a microphone while presenting at an ACM event, wearing a black blazer and glasses against a blue digital-themed background.

A former University of Maryland graduate student has received the 2026 Association for Computing Machinery (ACM) SIGMOBILE Dissertation Award for developing intelligent, energy-efficient sensing technologies that operate in highly resource-constrained environments by leveraging the physical properties of sound, radio and light.

Nakul Garg, who completed his Ph.D. in spring 2025 and is now an assistant professor of electrical and computer engineering at Rice University, earned the honor for his dissertation, “Toward Integrating Intelligence into Everything Around Us.”

The dissertation explored how intelligent technologies can operate with dramatically lower power requirements by redesigning sensing and computing architectures from the ground up.

At the center of the work is a physics-integrated computing paradigm that embeds portions of sensing and inference directly into physical processes rather than treating sensing, communication and computation as separate stages. By combining hardware design with algorithmic intelligence, the research enables sophisticated sensing and localization capabilities while consuming only a fraction of the energy required by conventional approaches.

“This work started from a simple question: How do we make intelligence available in places where traditional computing platforms simply cannot operate?” Garg said. “The vision of ambient intelligence has always been compelling, but existing systems remain too power-hungry and expensive to scale broadly. I wanted to rethink these architectures from first principles and explore how the physical world itself could participate in computation. It’s incredibly meaningful to see that work recognized by the SIGMOBILE community.”

Across acoustic, wireless and optical domains, Garg developed technologies that dramatically reduce hardware complexity while preserving performance. In acoustic sensing, his research demonstrated that specially engineered metamaterial structures could enable spatial audio perception using a single microphone instead of conventional multi-microphone arrays. In wireless systems, he introduced self-localization techniques that extract spatial information without synchronized antenna arrays, enabling compact, sticker-sized tracking devices capable of operating for years on minimal power budgets.

The dissertation also explored infrastructure-free coordination systems and non-invasive sensing techniques for applications such as monitoring food quality throughout global supply chains, demonstrating how these approaches could be deployed in real-world environments.

Garg’s adviser at UMD, Nirupam Roy, an associate professor of computer science with an appointment in the University of Maryland Institute for Advanced Computer Studies (UMIACS), said the work points toward a future in which intelligence is deeply integrated into the physical environment rather than confined to traditional computing devices.

“Nakul’s dissertation fundamentally changes how we think about intelligent sensing systems,” said Roy, who is a core faculty member in the Maryland Cybersecurity Center (MC2). “He showed that achieving powerful AI capabilities does not necessarily require larger models or more computation. By tightly integrating sensing, physical processes and inference, his work demonstrates a new class of technologies that are efficient, scalable and continuously aware of their environment. This research lays important groundwork for the future of pervasive and embedded AI.”

The SIGMOBILE Dissertation Award caps an exceptional period of innovation for Garg during his time at UMD. As a graduate researcher in Roy’s iCoSMoS Lab, he helped advance a wide range of next-generation information and communications technologies, earning the prestigious Marconi Society Paul Baran Young Scholar Award in 2024.

Among the notable projects he contributed to was “Owlet,” an ultra-low-power sound localization system for micro-robots that received the Best Demo Award at ACM MobiSys 2021. Garg also co-developed “ SING,” a novel system architecture presented at ICML 2025 that incorporates spatial speech understanding into LLMs, enabling machines to seamlessly understand the physical context and locations of speakers during conversations, much like humans do.

—Story by UMIACS communications group

Published May 28, 2026