Jack Collins

Art Institute of Portland, University of Oxford

Papers

Total Citations

H-Index

About

Jack Collins is an emerging robotics researcher whose work sits at the intersection of motion planning, reinforcement learning, and sim-to-real transfer — areas critical to deploying intelligent robots in real-world environments. His most cited work, "Leveraging Scene Embeddings for Gradient-Based Motion Planning in Latent Space" (2023, 7 citations), demonstrates how framing motion planning as optimisation within structured latent spaces can match traditional planning methods in success rates while dramatically improving computational speed, addressing a longstanding bottleneck in practical robotics. His 2024 paper, "TWIST: Teacher-Student World Model Distillation for Efficient Sim-to-Real Transfer" (5 citations), tackles the persistent sim-to-real gap in vision-based model-based reinforcement learning, offering a distillation framework that improves sample efficiency and real-world generalization. Rounding out his portfolio, "Efficient Skill Acquisition for Complex Manipulation Tasks in Obstructed Environments" (2023, 4 citations) addresses data-efficient robotic learning for small-batch assembly settings, enabling robust obstacle avoidance from minimal demonstrations. Collectively, Collins' research advances a coherent vision: making robot learning faster, more transferable, and practically deployable — contributions of growing relevance as robotics moves from controlled labs into complex, unstructured environments.