The number of everyday smart devices is projected to grow to billions in the coming decade. Our vision and mission are to make these intelligent networked devices more pervasive/ubiquitous and autonomous. Our research focuses on acquiring, analyzing, and augmenting ambient information with limited/less resources.

[Pervasive Perception] To effectively measure and monitor people's physical and behavioral health, continual and user-imperceptible sensing is essential. We focus on contact-based sensing by converting everyday objects into sensors while preserving their regular usability (Surface as Sensors).

• Vibration of Everything

• Structural Intelligence for Physical Signal Augmentation

[Autonomous Learning Adaptation] Due to the complexity of the physical world, sensing quality and data distributions can change significantly under different physical conditions, resulting in inaccurate detection and inference. We focus on autonomous assessing, configuring, and adapting heterogeneous sensing systems via cross-modal association using people as the shared context.

• Multi-Modal Collaborative Learning for Internet of Things (IoT)

• LLM-Enhanced Sensing Intelligence

[Networked Spatial Intelligence] Data from multimodal networked devices provides complementary information. By exploring the complementary and correlation of multimodal sensor data with different spatial/mobile characteristics (static, semi-mobile, mobile), we establish efficient inference/model transfer schemes guided by the modality's spatial characteristics.

• Artificial Intelligence of Things (AIoT) to Mitigate Socioeconomic Bias in Air Quality Monitoring

• Data-Driven Spatial Sensing Quality Assessment

UC Merced is located between Silicon Valley and the beautiful Yosemite National Park
#18 in Mobile Computing and #63 in general by CSRankings
#60 in U.S. News College Ranking
Established in 2005 and achieved R1 in 2025

I'm looking for highly motivated PhD students. 

Please contact me via email (span24@ucmerced.edu)

Sponsors

Alumni/Student Spin-off