 Adaptive Reference IOC, Arioc, is a new method for optimizing trajectories during natural walking. It uses a combination of direct collocation and gradient-based optimization to achieve faster convergence than other derivative-free methods. Arioc was tested on 10 subjects walking at different speeds and stride lengths, and it successfully reproduced the reference trajectory. This method could be used to optimize trajectories for assistive robotic systems like lower limb exoskeletons. This article was authored by Jiachen Wang, Eson Hashemi, and Arash Arami.