Sim2Real robot learning is an innovative field that aims to train robots in simulated environments and transfer their acquired skills to real-world applications. By combining simulation, machine learning, and robotics, Sim2Real robot learning tackles the challenges associated with training in real-world scenarios. This comprehensive report explores the core concepts and methodologies of Sim2Real robot learning, covering essential topics such as reinforcement learning, evolutionary robotics, simulators, domain adaptation, domain randomization, and meta learning. It delves into the advantages and limitations of simulation-based training, addressing the reality gap between simulations and real-world environments. The report highlights the importance of domain adaptation and domain randomization techniques in minimizing the discrepancies between simulated and real-world performance.