Edward Groshev

Papers

Total Citations

H-Index

About

Edward Groshev is a robotics researcher whose work tackles the fundamental challenge of enabling autonomous systems to perform complex, long-horizon tasks under uncertainty. His key research areas lie at the intersection of task and motion planning, reinforcement learning, and trajectory optimization. Groshev’s most impactful contribution is the **Interfaced Belief Space Planning (IBSP)** framework, introduced in his 2015 paper (42 citations), which provides a modular approach to integrating high-level task planning with low-level motion planning while explicitly reasoning about uncertainty. This work addresses a critical bottleneck in robotics: executing extended sequences of actions in partially observable environments. More recently, Groshev has developed the **Sub-Goal Trees** framework, a novel goal-based approach to reinforcement learning and trajectory prediction. This framework reformulates goal-directed problems by structuring trajectories around sub-goals, enabling more efficient learning and optimization. His 2020 paper (8 citations) extends this idea to multi-goal RL, while the 2019 work (6 citations) applies it to trajectory prediction. Groshev’s contributions are particularly notable for their theoretical elegance and practical potential, offering scalable solutions for robots operating in real-world, uncertain environments.