We present MotionDisco, a framework that discovers contact-rich, long-horizon humanoid loco-manipulation motions from scratch, without relying on teleoperation or motion retargeting from human demonstrations. This is challenging because the space of possible contact interactions grows combinatorially with the task horizon and the number of objects in the scene. MotionDisco enables rapid discovery of novel motions by coupling a large language model (LLM) guided evolutionary search over sequences of interactions with an efficient sequential kinodynamic trajectory optimizer and pruning strategy. Through extensive ablation studies, we show that our LLM-guided search discovers successful whole-body trajectories across several challenging long-horizon tasks. Finally, by training reinforcement learning tracking policies on the discovered trajectories, we transfer the motions to a real humanoid robot. This is the first work to discover and deploy long-horizon humanoid loco-manipulation skills entirely through automated evolutionary search.
@article{taouil26motiondisco,author={Taouil*, Ilyass and Ciebielski*, Michal and Omar*, Shafeef and Zhao, Haizhou and Dai, Angela and Johnson, Aaron M. and Khadiv, Majid},title={MotionDisco: Motion Discovery for Extreme Humanoid Loco-Manipulation},journal={arXiv preprint arXiv:2606.06139},year={2026},}
DynaRetarget: Dynamically-Feasible Retargeting using Sampling-Based Trajectory Optimization
Victor Dhedin*, Ilyass Taouil*, Shafeef Omar*, Dian Yu, Kun Tao, Angela Dai, and Majid Khadiv
DynaRetarget: Dynamically-Feasible Retargeting using Sampling-Based Trajectory Optimization.
@article{dhedin26dynaretarget,author={Dhedin*, Victor and Taouil*, Ilyass and Omar*, Shafeef and Yu, Dian and Tao, Kun and Dai, Angela and Khadiv, Majid},title={DynaRetarget: Dynamically-Feasible Retargeting using Sampling-Based Trajectory Optimization},journal={arXiv preprint arXiv:2602.06827},year={2026},}
Learning to Act Through Contact: A Unified View of Multi-Task Robot Learning
Learning to Act Through Contact: A Unified View of Multi-Task Robot Learning.
@inproceedings{omar25contact,author={Omar*, Shafeef and Khadiv, Majid},title={Learning to Act Through Contact: A Unified View of Multi-Task Robot Learning},booktitle={Proceedings of The 8th Annual Learning for Dynamics \& Control Conference (L4DC)},series={Proceedings of Machine Learning Research},publisher={PMLR},year={2026},}
2023
SafeSteps: Learning Safer Footstep Planning Policies for Legged Robots via Model-Based Priors
We present a footstep planning policy for quadrupedal locomotion that is able to directly take into consideration a-priori safety information in its decisions. At its core, a learning process analyzes terrain patches, classifying each landing location by its kinematic feasibility, shin collision, and terrain roughness. This information is then encoded into a small vector representation and passed as an additional state to the footstep planning policy, which furthermore proposes only safe footstep location by applying a masked variant of the Proximal Policy Optimization (PPO) algorithm [1]. The performance of the proposed approach is shown by comparative simulations on an electric quadruped robot walking in different rough terrain scenarios. We show that violations of the above safety conditions are greatly reduced both during training and the successive deployment of the policy, resulting in an inherently safer footstep planner. Furthermore, we show how, as a byproduct, fewer reward terms are needed to shape the behavior of the policy, which in return is able to achieve both better final performances and sample efficiency.
@inproceedings{omar23humanoids,author={Omar*, Shafeef and Amatucci, Lorenzo and Turrisi, Giulio and Barasuol, Victor and Semini, Claudio},title={SafeSteps: Learning Safer Footstep Planning Policies for Legged Robots via Model-Based Priors},booktitle={IEEE-RAS International Conference on Humanoid Robots},year={2023},}
2022
Fast Convex Visual Foothold Adaptation for Quadrupedal Locomotion
@inproceedings{omar22irim,author={Omar*, Shafeef and Amtucci, Lorenzo and Turrisi, Giulio and Barasuol, Victor and Semini, Claudio},booktitle={I-RIM Conference},title={Fast Convex Visual Foothold Adaptation for Quadrupedal Locomotion},year={2022},pages={143-146},doi={110.5281/zenodo.7531324},}
