Kamal Mokhtar
Papers
1
Total Citations
12
H-Index
1
About
Kamal Mokhtar is a rising force in robotic manipulation, specializing in deep reinforcement learning for complex, unstructured environments. His primary research focuses on developing intelligent policies that enable robots to autonomously interact with and rearrange highly cluttered spaces. Mokhtar’s most significant contribution is his pioneering work on self-supervised learning for joint pushing and grasping, where he demonstrated that a single, unified DRL policy can outperform traditional sequential approaches. His 2024 paper on this topic, already garnering 12 citations, presents a method that allows robots to dynamically decide when to push obstacles aside versus directly grasp a target, achieving robust performance in untrained, chaotic settings. This work directly addresses a critical bottleneck in industrial and service robotics: reliable object retrieval from bins or shelves. By eliminating the need for explicit environment modeling, Mokhtar’s approach marks a step toward truly adaptive, general-purpose manipulation. His research holds promise for applications in warehouse automation, assistive robotics, and household service, positioning him as a key innovator in the next generation of dexterous robotic systems.
Research Focus
Key Achievements
Top Papers
- 1