VLM-Grounded Task and Motion Planning With Uncertainty Aware Active Perception
Austine Oloo, Shiqi Zhang
Binghamton University/Computer Science
I am a Computer Science PhD student specializing in Embodied AI. My research bridges high-level symbolic reasoning and low-level sensorimotor control to enable intuitive Human-Robot Interaction — creating robotic systems that collaborate safely and naturally with people. I actively explore how modern foundation models can be grounded in robotic systems as powerful interfaces for perception and action.
My work sits at the intersection of robotics, computer vision, and AI — with a particular emphasis on their applications in the context of mobile manipulators.
Empowering Robots to Understand the Real World
Developing methods that enable robots to build rich, semantic understanding of 3D environments — fusing geometric perception with high-level reasoning so that mobile manipulators can interpret scenes, localize objects, and plan actions grounded in spatial awareness.
Enabling Robots to Transform the Real World
Building frameworks for robots to acquire, compose, and generalize manipulation skills — bridging symbolic task planning with learned sensorimotor policies so that mobile manipulators can perform complex, multi-step tasks in unstructured environments.
Austine Oloo, Shiqi Zhang
Austine Oloo*, Xinwei Guo*, Shiqi Zhang
Submitted to IROS 2026: Robot Planning and Situation Handling with Active Perception
Teaching Assistant for Intelligent Mobile Robotics
Successfully defended my Master’s thesis
Teaching Assistant for Introduction to AI
Started PhD in Computer Science at Binghamton University, advised by Prof. Shiqi Zhang
Successfully demonstrated autonomous robot navigation system in complex indoor environments
Started MSc in Computer Science at Binghamton University