Principal Applied AI Research Scientist

Abduallah
Mohamed, PhD

Principal Applied Research Scientist & technical lead at Meta Reality Labs, driving R&D across motion tracking, sensor fusion, and on-device AI for Smart Glasses and AR/VR. Core expertise spans IMU & multi-modal sensor fusion, indoor/outdoor motion tracking, SLAM & non-visual localization, multi-agent trajectory prediction, and LLM-based agentic frameworks. PhD, Electrical & Computer Engineering — UT Austin.

Open to collaboration — I'm currently co-supervising and collaborating with PhD students and researchers working on motion & agentic AI problems. Please reach out if you seek collaboration.

scroll

Areas of Expertise

Applied AI

Real-time ML for edge devices — fitness tracking, AR/VR, sensor fusion, trajectory analysis with strict memory and power constraints.

Edge MLSensor FusionFitness AIAR/VR

Motion & Trajectory

Deep learning models for multi-agent motion forecasting, SLAM, non-visual localization, and multi-level data fusion systems.

Trajectory PredictionSLAMLocalizationGraph CNNs

Sensors

Full-stack sensor system development from signal quality assessment to application abstraction. IMU, GNSS, magnetometers, ultrasonic.

IMUGNSSMagnetometersUltrasonic

Agentic Systems

LLM-driven swarm intelligence frameworks, automated hardware debugging pipelines, and autonomous engineer persona agents.

LLM AgentsSwarm AIAuto-debuggingPipelines

Experience

Principal Applied Research Scientist

Meta Reality Labs
2022 – PresentRedmond, WA

Technical lead across 4+ product groups and 6+ teams. Research in outdoor/indoor motion tracking, fitness AI, SLAM pipelines, and non-visual localization for Smart Glasses and AR/VR devices. Architected LLM-based swarm intelligence frameworks for product development.

SLAMSensor FusionEdge AILLM AgentsAR/VR
Meta Reality Labs
PhD InternshipsMeta · Apple · Intuitive SurgicalDuring PhD at UT Austin2018 – 2021Research internships at Meta, Apple, and Intuitive Surgical during doctoral studies at UT Austin.

R&D Autonomous Driving Algorithms

Avelabs
2016 – 2017Egypt

Technical lead for the algorithms team. Developed real-time deep learning systems for object tracking, lane detection, road recognition, and traffic sign recognition for autonomous vehicles.

Object DetectionLane DetectionDeep LearningAutonomous Driving

R&D Autonomous Driving Algorithms

Valeo
2015 – 2016Egypt

Designed algorithm for processing raw ultrasonic sensor data to detect and track multiple objects using unsupervised deep learning — subsequently patented. Worked on static mapping using Extended Kalman Filters.

Ultrasonic SensorsEKFUnsupervised LearningPatent

Software Engineer

ITWorx
2014 – 2015Egypt

Designed, executed, assessed, and troubleshot software programs and web-based applications across multiple client projects.

Software EngineeringWeb Applications

Co-founder & ML Consultant

AR/VR Startup · Freelance
2011 – 2014Egypt

Co-founded and ran a startup in the AR/VR space. Simultaneously worked as a freelance machine learning consultant specializing in computer vision.

AR/VRComputer VisionEntrepreneurshipFreelance ML

Publications

1,699+ citations across CVPR, ECCV, ICCV, IEEE PerCom, and WACV. Google Scholar

2026
iMotion-LLM: Instruction-Conditioned Trajectory Generation
Proceedings of the IEEE/CVF Winter Conference on Applications of Computer …, 2026
2024
A review of deep learning for video captioning
IEEE Transactions on Pattern Analysis and Machine Intelligence, 2024 63 citations
2022
HAR-GCNN: Deep graph CNNs for human activity recognition from highly unlabeled mobile sensor data
2022 IEEE International Conference on Pervasive Computing and Communications …, 2022 42 citations
2022
2021
2020
2020
Social-stgcnn: A social spatio-temporal graph convolutional neural network for human trajectory prediction
Proceedings of the IEEE/CVF conference on computer vision and pattern …, 2020 1334 citations
2018
An IMU-based traffic and road condition monitoring system
HardwareX 4, e00045, 2018 37 citations
2018
MCRM: Mother compact recurrent memory
arXiv preprint arXiv:1808.02016, 2018 1 citations
Education

PhD, Electrical & Computer Engineering

The University of Texas at Austin · 2017–2022
Supervised by Prof. Claudel & Prof. Tewfik · Co-advised by Prof. Elhoseiny (KAUST/Stanford)
Research Focus

Trajectory prediction for autonomous and non-autonomous objects. Member of the Mobile Automation and Sensing Systems (MASS) lab.

Contact

Open to research collaborations, speaking engagements, and advisory opportunities in applied AI, sensor systems, and autonomous systems.