About

Omid Abari is an Associate Professor at UCLA Samueli School of Engineering, specializing in computer science and electrical engineering. His research interests include Internet-of-Things (IoT), wireless networks, mobile systems, software-hardware systems, and human-computer interaction (HCI). He holds a PhD and an MS from the Massachusetts Institute of Technology and a BEng from Carleton University. Recognized for his contributions, he received the NSF CAREER Award in 2023. His work focuses on advancing technologies in wireless communication and human-computer interfaces, contributing to the development of innovative solutions in these fields.

Research topics

• Computer Science
• Engineering
• Computer network
• Telecommunications
• Agronomy
• Remote sensing
• Environmental science
• Meteorology
• Operating system
• Real-time computing
• Electrical engineering
• Agricultural engineering
• Computer hardware
• Embedded system
• Soil science
• Geotechnical engineering

Selected publications

• Deform to Inform: Persistent Batteryless Sensing via Antenna Deformation and RFID Impedance Adaptation

  2026-05-08

  articleOpen accessSenior author

  A stable environment is critical for preserving product quality in industries such as pharmaceuticals and perishable goods logistics. While current RFID-based sensors enable wireless monitoring of temperature and moisture, they lack the ability to record threshold violations without continuous reader power or battery-supported logging. This paper presents AntSense, a battery-free RFID sensor that leverages stimuli-responsive structures to physically record environment violation events during the transit even when there is no reader available. When exposed to temperature or moisture beyond thresholds, memory alloy or solvable materials undergo irreversible geometric deformation, altering the integrated antenna’s characteristics. These changes remain detectable via standard RFID readers even after environment conditions normalize, enabling verification of transit integrity. Our prototype evaluations demonstrate reliable detection of threshold violations in real-world scenarios. By combining passive operation with persistent event recording, AntSense offers a scalable solution for supply chain monitoring without compromising the maintenance-free advantages of RFID technology.

  PublisherDOI

• XR Devices Send WiFi Packets When They Should Not: Cross-Building Keylogging Attacks via Non-Cooperative Wireless Sensing

  2026-01-01

  articleOpen accessSenior author

  3) Signal Processing 10s of meters

  PublisherDOI

• Camera-RFID Fusion for Robust Asset Tracking in Forested Environments

  arXiv (Cornell University) · 2026-04-29

  preprintOpen accessSenior author

  Passive RFID tags offer a cost-effective and scalable solution for tracking numerous deployed assets. However, in forested environments, signal attenuation and multipath effects generally limit RFID spatial accuracy to the meter level. Conversely, while cameras employing stereo vision can achieve centimeter-level precision, relying solely on computer vision fails to resolve issues arising from spatial association ambiguity and partial occlusions in dense settings. Fusing these modalities allows systems to harness the high-accuracy benefits of vision while retaining the robust, non-line-of-sight identification advantages of RFID. Yet, a primary challenge in achieving this, which is the central focus of this paper, lies in accurately associating the disparate trajectories generated by these two sensors. To overcome this limitation, we introduce a novel camera--RFID fusion framework that integrates depth and object information with advanced trajectory-matching algorithms. By successfully bridging the meter-to-centimeter accuracy gap, the proposed approach helps achieve reliable tag localization even when assets temporarily leave the camera's field of view. To the best of our knowledge, this represents the first application of camera--RFID fusion for asset tracking in natural forested environments.

  PublisherDOI

• Camera-RFID Fusion for Robust Asset Tracking in Forested Environments

  ArXiv.org · 2026-04-29

  articleOpen accessSenior author

  Passive RFID tags offer a cost-effective and scalable solution for tracking numerous deployed assets. However, in forested environments, signal attenuation and multipath effects generally limit RFID spatial accuracy to the meter level. Conversely, while cameras employing stereo vision can achieve centimeter-level precision, relying solely on computer vision fails to resolve issues arising from spatial association ambiguity and partial occlusions in dense settings. Fusing these modalities allows systems to harness the high-accuracy benefits of vision while retaining the robust, non-line-of-sight identification advantages of RFID. Yet, a primary challenge in achieving this, which is the central focus of this paper, lies in accurately associating the disparate trajectories generated by these two sensors. To overcome this limitation, we introduce a novel camera--RFID fusion framework that integrates depth and object information with advanced trajectory-matching algorithms. By successfully bridging the meter-to-centimeter accuracy gap, the proposed approach helps achieve reliable tag localization even when assets temporarily leave the camera's field of view. To the best of our knowledge, this represents the first application of camera--RFID fusion for asset tracking in natural forested environments.

  PublisherOA PDF

• FruitScope: A Non-Invasive Fruit Ripeness Sensing System via Multi-Resolution FMCW Design and Acoustic Sensing

  2026-05-08

  articleOpen accessSenior author
  PublisherDOI

• Motion Capture with Millimeter-Wave Tags

  2026-05-08

  articleOpen access

  This paper introduces M3oCap, a millimeter-wave (mmWave) tag-based motion capture system that delivers accurate 6 degrees of freedom motion tracking. M3oCap utilizes a single commercial mmWave radar and custom-designed mmWave backscatter tags (2.5 cm × 3 cm) to localize and track the motion of tagged objects. Our system features novel algorithms that effectively isolate weak backscattered signals while accurately recovering phase changes induced by motion, allowing high-rate tracking of tag movements. Experiments show that M3oCap achieves performance close to commercial motion capture systems – it provides 1000 measurements per second with a median range tracking accuracy of 250 μ m, median localization accuracies of 3.34 mm, 3.95 mm, and 4.20 mm in x, y, z dimensions, and median orientation tracking accuracies of 1.7°, 2.3°, and 2.0° across pitch, roll, and yaw. We further demonstrate the system’s fine-grained movement tracking capabilities through various motion capture tasks.

  PublisherDOI

• Tracking Wildfire Assets with Commodity RFID and Gaussian Process Modeling

  arXiv (Cornell University) · 2025-12-17

  preprintOpen accessSenior author

  This paper presents a novel, cost-effective, and scalable approach to track numerous assets distributed in forested environments using commodity Radio Frequency Identification (RFID) targeting wildfire response applications. Commodity RFID systems suffer from poor tag localization when dispersed in forested environments due to signal attenuation, multi-path effects and environmental variability. Current methods to address this issue via fingerprinting rely on dispersing tags at known locations {\em a priori}. In this paper, we address the case when it is not possible to tag known locations and show that it is possible to localize tags to accuracies comparable to global positioning systems (GPS) without such a constraint. For this, we propose Gaussian Process to model various environments solely based on RF signal response signatures and without the aid of additional sensors such as global positioning GPS or cameras, and match an unknown RF to the closest match in a model dictionary. We utilize a new weighted log-likelihood method to associate an unknown environment with the closest environment in a dictionary of previously modeled environments, which is a crucial step in being able to use our approach. Our results show that it is possible to achieve localization accuracies of the order of GPS, but with passive commodity RFID, which will allow the tracking of dozens of wildfire assets within the vicinity of mobile readers at-a-time simultaneously, does not require known positions to be tagged {\em a priori}, and can achieve localization at a fraction of the cost compared to GPS.

  PublisherDOI

• OmniVLA: Physically-Grounded Multimodal VLA with Unified Multi-Sensor Perception for Robotic Manipulation

  arXiv (Cornell University) · 2025-11-03

  preprintOpen access

  Vision-language-action (VLA) models have shown strong generalization for robotic action prediction through large-scale vision-language pretraining. However, most existing models rely solely on RGB cameras, limiting their perception and, consequently, manipulation capabilities. We present OmniVLA, an omni-modality VLA model that integrates novel sensing modalities for physically-grounded spatial intelligence beyond RGB perception. The core of our approach is the sensor-masked image, a unified representation that overlays spatially grounded and physically meaningful masks onto the RGB images, derived from sensors including an infrared camera, a mmWave radar, and a microphone array. This image-native unification keeps sensor input close to RGB statistics to facilitate training, provides a uniform interface across sensor hardware, and enables data-efficient learning with lightweight per-sensor projectors. Built on this, we present a multisensory vision-language-action model architecture and train the model based on an RGB-pretrained VLA backbone. We evaluate OmniVLA on challenging real-world tasks where sensor-modality perception guides the robotic manipulation. OmniVLA achieves an average task success rate of 84%, significantly outperforms both RGB-only and raw-sensor-input baseline models by 59% and 28% respectively, meanwhile showing higher learning efficiency and stronger generalization capability.

  PublisherOA PDFDOI

• Tracking Wildfire Assets With Commodity RFID and Gaussian Process Modeling

  IEEE Journal of Radio Frequency Identification · 2025-01-01

  articleOpen accessSenior author

  This paper presents a novel, cost-effective, and scalable approach to track numerous assets distributed in forested environments using commodity Radio Frequency Identification (RFID) targeting wildfire response applications. Commodity RFID systems suffer from poor tag localization when dispersed in forested environments due to signal attenuation, multi-path effects and environmental variability. Current methods to address this issue via fingerprinting rely on dispersing tags at known locations <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">a priori</i>. In this paper, we address the case when it is not possible to tag known locations and show that it is possible to localize tags to accuracies comparable to global positioning systems (GPS) without such a constraint. For this, we propose Gaussian Process to model various environments solely based on RF signal response signatures and without the aid of additional sensors such as global positioning GPS or cameras, and match an unknown RF to the closest match in a model dictionary. We utilize a new weighted log-likelihood method to associate an unknown environment with the closest environment in a dictionary of previously modeled environments, which is a crucial step in being able to use our approach. Our results show that it is possible to achieve localization accuracies of the order of GPS, but with passive commodity RFID, which will allow the tracking of dozens of wildfire assets within the vicinity of mobile readers at-a-time simultaneously, does not require known positions to be tagged <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">a priori</i>, and can achieve localization at a fraction of the cost compared to GPS.

  PublisherOA PDFDOI

• Sustainable and Low-Cost Greenhouse Soil Moisture Monitoring Using Battery-Free RFID Sensors

  ACM Transactions on Sensor Networks · 2025-01-24 · 3 citations

  article

  Intelligent irrigation based on measurements of soil moisture levels in every pot in a greenhouse can not only improve plant productivity and quality but also save water. However, existing soil moisture sensors are too expensive to deploy in every pot. We therefore introduce GreenTag, a low-cost RFID-based soil moisture sensing system whose accuracy is comparable to that of an expensive soil moisture sensor. Our key idea is to attach two RFID tags to a plant’s container so that changes in soil moisture content are reflected in their Differential Minimum Response Threshold (DMRT) metric at the reader. We show that a low-pass filtered DMRT metric is robust to changes both in the RF environment (e.g., from human movement) and in pot locations. In addition, we propose a fast DMRT acquisition algorithm and a time-efficient tag query protocol, which can reduce the sensing latency by 90%. In a realistic setting, GreenTag achieves a 90-percentile moisture estimation errors of 5%, which is comparable to the 4% errors using expensive soil moisture sensors. Moreover, this accuracy is maintained despite changes in the RF environment and container locations. We also show the effectiveness of GreenTag in a real greenhouse.

  PublisherDOI

Frequent coauthors

• Dina Katabi

  Massachusetts Institute of Technology

  19 shared

• Mohammad Hossein Mazaheri

  University of California, Los Angeles

  17 shared

• Ali Abedi

  University of Tabriz

  13 shared

• Haitham Hassanieh

  9 shared

• Srinivasan Keshav

  University of Cambridge

  8 shared

• Tim Brecht

  8 shared

• Haofan Lu

  7 shared

• Ali Abedi

  Berkeley College

  6 shared

Awards & honors

• NSF CAREER Award (2023)