I am a Ph.D. candidate in Computer Science at George Mason University and a member of the Robotic Anticipatory Intelligence & Learning (RAIL) Lab, advised by Dr. Greg Stein. My research sits at the intersection of robotics, machine learning, and task and motion planning, with a focus on anticipatory decision-making for household and service robots.
More broadly, my research focuses on enabling robots to reason about the consequences of their actions and guiding them toward forward-thinking behaviors such as organization, tidiness, and preparation. This opens exciting opportunities to develop robots capable of serving as long-term assistive companions.
Or if you're interested in a famous Nepali board game (multiplayer or play against AI opponent), I built बाघचाल (Bagh Chal) — click here to play!
My work focuses on enabling long-lived autonomy for robots and embodied AI operating in persistent, real-world environments. I study how intelligent agents can anticipate the impact of their actions and make better decisions in the present that would otherwise be difficult to achieve. My research spans anticipatory task and motion planning, learning-augmented planning, and lifelong robot autonomy. Most recently, I have developed methods that integrate anticipatory decision-making estimates into classical planning systems, guiding search toward world states that improve efficiency, robustness, and performance across sequences of future tasks.
Selected recent updates.
Recent work in anticipatory planning, task and motion planning, and graph learning.
We present a learning-augmented task and motion planning framework that anticipates the future impact of current actions and selects plans that reduce long-term cost across sequences of rearrangement tasks in persistent continuous-space environments.
We introduce anticipatory task planning, a formulation that helps robots solve current tasks while reasoning about the expected cost of unseen future tasks.
We model floor plans as graphs and use graph neural networks to improve room classification performance by leveraging relational structure in building layouts.
Selected technical projects beyond published papers.
I built a system that injects learned anticipatory cost models into the Fast Downward planning stack, allowing neural cost estimation to be composed with standard symbolic heuristics such as FF, max, and goal-count.
In a 2D navigation setting, I formulated the domain in PDDL and used PDDLStream to plan obstacle relocations that open a feasible path to the goal.
I compared Proximal Policy Optimization (PPO) and Deep Q-Networks (DQN) on the Lunar Lander benchmark to study differences in learning stability, sample efficiency, and control performance.
A famous Nepali strategy board game playable in single-player or local multiplayer mode. In single-player, challenge different AI opponents — a built-in minimax engine, Llama/GPT via Pollinations (free), Google Gemini Flash (free tier), or my own AlphaGo-inspired neural network trained on Bagh Chal self-play. Built with assistance from Claude.