Hamid Hadjar

Papers

Total Citations

H-Index

About

Hamid Hadjar is a roboticist advancing the frontier of contact-rich manipulation through sim-to-real learning. His research focuses on bridging the reality gap—the discrepancy between simulated and physical environments—to enable robots to perform complex assembly tasks without extensive real-world training. His most-cited work, "A Sim-to-Real Learning-Based Framework for Contact-Rich Assembly by Utilizing CycleGAN and Force Control" (2023, 28 citations), introduces a novel framework that combines CycleGAN-based domain adaptation with force control to transfer policies trained in simulation to real robots. This approach significantly reduces the need for costly real-world data collection while maintaining robust performance in high-precision assembly. Hadjar’s contributions address a critical bottleneck in deep reinforcement learning for robotics: the safety and sample efficiency challenges of training directly in physical environments. By leveraging generative models to align simulated and real sensory inputs, his work demonstrates a practical path toward deploying learned policies in industrial settings. His research holds promise for automating manufacturing tasks that require delicate force modulation, such as peg-in-hole insertion and gear assembly.