http://localhost:1313/tags/virtual-fixtures/ Recent content in Virtual-Fixtures on Markus Knauer Hugo en-us Fri, 24 Apr 2026 00:00:00 +0000 http://localhost:1313/posts/momo-physical-verbal-graphical-skill-adaptation/ Fri, 24 Apr 2026 00:00:00 +0000 http://localhost:1313/posts/momo-physical-verbal-graphical-skill-adaptation/ <p>What if non-expert operators could adapt an industrial robot&rsquo;s skills as easily as touching, talking to, or clicking on it? Our framework <strong>MOMO</strong> (<strong>Mo</strong>tion <strong>Mo</strong>dulation) unifies three complementary interaction modalities for seamless robot skill learning and adaptation.</p> <picture class=""> <source srcset="http://localhost:1313/images/momo_overview.webp" type="image/webp"> <img src="http://localhost:1313/images/momo_overview.png" alt="Overview of the MOMO framework unifying physical, verbal, and graphical modalities for robot skill adaptation" loading="lazy"> </picture> <p>MOMO integrates five components around a central motion modulation module:</p> <ul> <li><strong>Kinesthetic touch</strong> for precise spatial corrections, using energy-tank-based human intention detection that automatically inserts via-points into the underlying motion model.</li> <li><strong>Natural language</strong> for high-level semantic modifications through a tool-based LLM architecture that selects and parameterizes pre-validated functions — never generating executable code.</li> <li><strong>A graphical web interface</strong> (Vue.js/Three.js) for visualizing geometric relations, inspecting parameters, and editing via-points via drag-and-drop on a real-time digital twin.</li> <li><strong>Probabilistic Virtual Fixtures</strong> for guided demonstration recording.</li> <li><strong>Ergodic control</strong> for surface finishing tasks.</li> </ul> <p>A key result: the tool-based LLM architecture <strong>generalizes beyond KMP-based skills to ergodic control</strong>, enabling the same chat interface to drive voice-commanded surface finishing. Users freely switch between modalities — voice for obstacle avoidance, kinesthetic for fine corrections, graphical for verification.</p>