Robotics researcher focused on navigation, embodied AI, and spatial intelligence.
I received my PhD from the Aerial Robotics Group at HKUST, advised by Prof. Shaojie Shen. My research spans robust localization, mapping, cross-view perception, and learning-based navigation.
Currently, I am a VLA Research Intern at China Merchants Lion Rock AI Lab, studying spatial generalization for robotic systems. Previously, I led mapping and localization algorithms at Unity Drive Innovation.
Paper "Do Embodied AI Models Have Physical Intuition? An Egocentric Probing Benchmark" (first author) accepted to the Tactile Sensing for Robotic Foundation Models Workshop at RSS 2026. Poster
Paper "The Moving Eye: Enhancing VLA Spatial Generalization via Hybrid Dynamic Data Collection" (co-first author) accepted to IROS 2026. Project page
Successfully defended my PhD thesis: "Robot Navigation from Explicit Geometry to Implicit Models". Grateful to my committee: Prof. Kun Xu (Chair), Prof. Shaojie Shen (Supervisor), Prof. Wei Zhang, Prof. Zili Meng, Prof. Yang Gao, and external examiner Prof. Dimitrios Kanoulas (UCL, UK).
Released my iOS app RTK Helper, a professional high-precision positioning tool on the App Store.
Attended Meta Research Summit in Seattle. Honored to meet Prof. Shuran Song, Prof. Danfei Xu, Prof. Xiaolong Wang and many other outstanding partners. Special thanks to my friends Jianhao Jiao and Yifu Tao for their support — hoping Aria accelerates the arrival of AGI!
Invited to serve as Session Co-Chair for "Autonomous Vehicles 3" at IROS 2025. Looking forward to facilitating discussions in Hangzhou, China.
Paper accepted to IROS 2025 on Visual Localization using Novel Satellite Imagery (first author, Oral). See you in Hangzhou, China.
Paper accepted to IEEE Transactions on Instrumentation and Measurement on Global Optimal Solutions to Scaled Quadratic Pose Estimation Problems.
Paper accepted to IEEE Transactions on Instrumentation and Measurement on Globally Optimal Estimation of Accelerometer-Magnetometer Misalignment.
Paper accepted to IEEE Robotics and Automation Letters on Efficient Camera Exposure Control for Visual Odometry via Deep Reinforcement Learning.
My research bridges classical robotics and learning-based methods, focusing on robust, generalizable navigation and manipulation systems. I am interested in augmenting physically grounded models with high-level semantic reasoning and generative priors, as well as improving the spatial generalization of Vision-Language-Action (VLA) models for embodied AI.
Do embodied AI models possess physical intuition from egocentric experience? We build a probing benchmark that captures fine-grained hand-object interactions across deformable, rigid, and fluid-filled objects, exposing where current vision-language models succeed and fail at predicting forces, compliance, and tactile outcomes from first-person video alone. [Project Page]
Enhancing VLA spatial generalization via hybrid dynamic data collection. By mixing multi-fixed and moving camera viewpoints, we break shortcut learning (camera-base and object-position coupling) and achieve robust manipulation under unseen camera poses and object configurations. The strategy generalizes across VLA architectures including ACT, Diffusion Policy, Pi0, and Gr00t. [Project Page]
Developing tightly-coupled optimization frameworks using LiDAR, IMU, and UWB to achieve drift-free, real-time state estimation.
Exploring BEV-based geometric and semantic alignment between ego-view and satellite-view images to enable cross-view localization.
A system integrating motion distortion correction, IMU bias estimation, and Kd-tree-accelerated scan matching. RLIL achieves centimeter-level accuracy in GNSS-denied environments.
Constructing static TSDF maps with dynamic object removal and using scan-to-map deviations to filter dynamic points in real-time.
Applying diffusion models to generate semantically aligned BEV representations from monocular images.
Integrates IPM, Navier-Stokes inpainting, and CLIP-based description to synthesize BEV representations for matching against satellite imagery.
Integrating fast reactive modules with slow semantic planners leveraging satellite imagery and VLMs to enable long-horizon, cross-modal Vision-Language Navigation and early-warning instructions.
Study notes on Vision-Language-Action (VLA) model architectures, hosted at vla.yilong-zhu.com.