Shaahin Angizi
New Jersey Institute of Technology, NJ·Engineering·Early-career
Smart devices such as cameras, wearables and environmental sensors gather huge amounts of data. Their small batteries and weak processors make it hard to run artificial intelligence (AI) on that data.
- Fields
Engineering
Computer Science
- Topics
in-sensor computing
processing-in-sensor
in-memory computing
- Grants
- 3 known grants, through Jul 2031
What they work on
- Design reconfigurable circuits and architectures for a multi-modal in-sensor accelerator that moves AI computation onto sensor nodes as efficiently as possible.
- Develop modality-sequencing software, energy-performance controllers and graph neural network predictors that plan how multi-sensor devices execute AI tasks.
- Build a large language model-driven platform, with structured datasets and automated pipelines, that generates in-sensor accelerator hardware designs automatically.
- Repurpose processor cache memory into a data-parallel processing unit so edge devices can run AI image processing near the sensor, paired with hardware-oriented algorithms that reduce memory access.
- Design a hybrid Processing-In-Sensor and Processing-Near-Sensor platform with adjustable-precision arithmetic for machine learning on continuously or energy-harvesting-powered imagers.
- Create IRISC, which saves intermediate results in non-volatile spin-based devices during power loss, and test it on FPGA-MRAM emulation platforms.
More about the research
Sending everything to the cloud costs energy and time. The projects here share one idea: do the computing right where the data is collected, inside or next to the sensor or in the device's own memory. The work spans circuits, chip architectures and algorithms for the Internet of Things (IoT), the growing web of connected devices.
Why it matters
Battery-limited devices for health monitoring, environmental and industrial sensing, cars, agriculture and security struggle to run AI locally. Faster, lower-power in-sensor and in-memory computing could let these devices analyze data on the spot. This would cut energy use and delays, and could keep devices powered by harvested energy running reliably. Open-source hardware cores, datasets and tools could also help chip and device designers.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
CAREER: Opportunistic and Adaptive Multi-Modal In-Sensor Computing for Artificial Intelligence of Things
NSF · Aug 2026 – Jul 2031
Collaborative Research: Integrated Sensing and Normally-off Computing for Edge Imaging Systems
NSF · Aug 2022 – Jan 2027
CNS Core: Small: Toward Opportunistic, Fast, and Robust In-Cache AI Acceleration at the Edge
NSF · Jan 2023 – Dec 2026
Closest labs
Mimi Xie
University of Texas at San Antonio, TX·Computer Science·Early-career
Redesign and optimize tiny machine learning models for battery-less, energy-harvesting IoT devices so that inference and decisions still happen on time under strict power, memory, and computation limits.
Deliang Fan
Arizona State University Campus Immersion, AZ·Engineering·Early-career
Design, fabricate and validate a hybrid in-memory computing chip that combines RRAM, a non-volatile resistive memory, with SRAM to run deep neural networks with high energy efficiency.
Yuhao Zhu
University of Rochester, NY·Computer Science·Early-career
Develop the foundations for in-sensor machine vision that combines computational metasurfaces (flat nanostructures that process light) with analog encoders inside CMOS image sensors, so images are partly processed before they are digitized.
The school
New Jersey Institute of Technology: 12 labs on PI Pages, $172M a year on research, 74 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.