# MOMO: A framework for seamless physical, verbal, and graphical robot skill learning and adaptation

arXiv: https://arxiv.org/abs/2604.20468 · HTML: https://mlsanity.com/paper/2604.20468/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 8 of its 33 references that are in the map.
- First posted on arXiv 22 Apr 2026.

## Builds on (most cited first, 8 of 8)

- [Toolformer: Language Models Can Teach Themselves to Use Tools](https://mlsanity.com/paper/2302.04761/) (2023)
- [Code as Policies: Language Model Programs for Embodied Control](https://mlsanity.com/paper/2209.07753/) (2022)
- [What Are Tools Anyway? A Survey from the Language Model Perspective](https://mlsanity.com/paper/2403.15452/) (2024)
- [Kernelized Movement Primitives](https://mlsanity.com/paper/1708.08638/) (2017)
- [Interactive incremental learning of generalizable skills with local trajectory modulation](https://mlsanity.com/paper/2409.05655/) (2024)
- [OVITA: Open-Vocabulary Interpretable Trajectory Adaptations](https://mlsanity.com/paper/2508.17260/) (2025)
- [A Unified Framework for Probabilistic Dynamic-, Trajectory- and Vision-based Virtual Fixtures](https://mlsanity.com/paper/2506.10239/) (2025)
- [IROSA: Interactive Robot Skill Adaptation using Natural Language](https://mlsanity.com/paper/2603.03897/) (2026)

## Abstract (arXiv)

Industrial robot applications require increasingly flexible systems that non-expert users can easily adapt for varying tasks and environments. However, different adaptations benefit from different interaction modalities. We present an interactive framework that enables robot skill adaptation through three complementary modalities: kinesthetic touch for precise spatial corrections, natural language for high-level semantic modifications, and a graphical web interface for visualizing geometric relations and trajectories, inspecting and adjusting parameters, and editing via-points by drag-and-drop. The framework integrates five components: energy-based human-intention detection, a tool-based LLM architecture (where the LLM selects and parameterizes predefined functions rather than generating code) for safe natural language adaptation, Kernelized Movement Primitives (KMPs) for motion encoding, probabilistic Virtual Fixtures for guided demonstration recording, and ergodic control for surface finishing. We demonstrate that this tool-based LLM architecture generalizes skill adaptation from KMPs to ergodic control, enabling voice-commanded surface finishing. Validation on a 7-DoF torque-controlled robot at the Automatica 2025 trade fair demonstrates the practical applicability of our approach in industrial settings.