Vaishak Belle

About

Vaishak Belle is a prominent AI researcher whose work sits at the intersection of knowledge representation, probabilistic reasoning, and cognitive robotics. Best known for his contributions to planning and reasoning under uncertainty, Belle has developed foundational frameworks for enabling intelligent agents to operate in partially observable, noisy, and dynamically changing environments. His highly cited work on contingent planning (48 citations) addresses one of AI's core challenges—how agents can reason and act when they cannot directly observe the world—with applications spanning robotics, game playing, and diagnostic systems. Through projects like PREGO and ALLEGRO, he has advanced belief-based programming languages that bridge high-level logical reasoning with the continuous probabilistic uncertainty inherent in real-world robotics. His theoretical contributions extend to the situation calculus, where he has significantly enriched its expressive power to handle continuous uncertainty and unbounded first-order dynamical domains. More recently, Belle has turned toward ethical AI, exploring how knowledge representation and machine learning can be aligned responsibly in high-stakes domains such as recidivism prediction and credit risk analysis (18 citations). With a body of work spanning over a decade and accumulating hundreds of citations, Belle has established himself as a leading voice in making AI systems both logically rigorous and practically grounded.

Research Focus

Key Achievements

Top Papers

1. 1
   Computing Contingent Plans via Fully Observable Non-Deterministic Planning

   48 citations · 2014
2. 2
   Reasoning about continuous uncertainty in the situation calculus

   22 citations · 2013
3. 3
   ALLEGRO: belief-based programming in stochastic dynamical domains

   20 citations · 2015
4. 4
   Knowledge representation and acquisition for ethical AI: challenges and opportunities

   18 citations · 2023
5. 5
   Reasoning about discrete and continuous noisy sensors and effectors in dynamical systems

   18 citations · 2018
6. 6
   Reasoning about Probabilities in Unbounded First-Order Dynamical Domains

   16 citations · 2017
7. 7
   PREGO: An Action Language for Belief-Based Cognitive Robotics in Continuous Domains

   15 citations · 2014
8. 8
   Epistemic planning: Perspectives on the special issue

   14 citations · 2022
9. 9
   Randomized trees for real-time one-step face detection and recognition

   14 citations · 2008
10. 10
    How to progress beliefs in continuous domains

    11 citations · 2014

Key Collaborators

Contact & Links