Research Scientist- Sensor Fusion & GNSS-Degraded Navigation

About the Role:

This is one of our client requirements

We are building a novel perception and localization system that can understand motion and spatial context even when GNSS signals are weak or unavailable. As a Research Scientist (Sensor Fusion & Inertial Navigation), you will lead algorithmic research at the intersection of physics-based modeling, estimation theory, and data-driven learning. Your work will focus on developing accurate, drift-resilient motion inference using inertial, magnetic, barometric, and environmental sensors all within the constraints of consumer-grade hardware. Youll collaborate closely with the AI/ML and systems teams to bridge the gap between classical navigation theory and modern machine learning.

Research Focus Areas:

• Inertial navigation and dead-reckoning under GNSS-degraded or denied conditions.
• Sensor fusion combining IMU, magnetometer, barometer, Wi-Fi/BLE, and GNSS.
• Drift modeling, bias correction, and error propagation analysis in MEMS sensors.
• Design of robust estimators Kalman, Extended/Unscented Kalman, Particle, Complementary, or hybrid physics + ML filters.
• Development of synthetic and real-world datasets for model benchmarking.
• Hybrid methods: integrating deep learning with physics-based state estimation.
• Contribution to patents and academic publications in localization, navigation, and perception.

Key Responsibilities:

• Lead research in sensor fusion algorithms for smartphone-grade or embedded sensors.
• Formulate mathematical models for bias, drift, and sensor error propagation.
• Prototype and validate new algorithms using Python, MATLAB, or C++. • Design experiments for indoor and semi-outdoor GNSS-degraded testing.
• Collaborate with ML engineers to integrate fusion outputs into backend inference pipelines.
• Write technical documentation, research notes, and white papers.
• Mentor engineers and guide graduate interns in research methodology.

Experience:

• Demonstrated work in sensor fusion, navigation, SLAM, or state estimation.
• Solid understanding of inertial navigation mathematics and Kalman filtering frameworks.
• Experience handling multi-sensor time-series (IMU, magnetometer, barometer).
• Strong proficiency in Python / MATLAB / C++ for algorithmic research. • Proven track record of publications, patents, or open-source contributions in related fields.

Preferred / Bonus Skills:

• Exposure to Deep Inertial Odometry, IONet, or RoNIN-type architectures.
• Familiarity with Allan variance, sensor calibration, and temperature-drift analysis.
• Knowledge of ROS, simulation environments, or motion-capture systems.
• Interest in combining model-based and data-driven methods for perception.
• Ability to design and interpret controlled experiments with consumer sensors