首页|New Findings from Center for Research and Advanced Studies in Robotics Provides New Insights (Assist-as-needed Robotic Strategy Based On Velocity Fields for Enh ancing Motor Training)
New Findings from Center for Research and Advanced Studies in Robotics Provides New Insights (Assist-as-needed Robotic Strategy Based On Velocity Fields for Enh ancing Motor Training)
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NETL
NSTL
New research on Robotics is the subjec t of a report. According to news reporting out of Saltillo, Mexico, by NewsRx ed itors, research stated, "Designing robotic assistance strategies that prioritize users' effort and minimize robot intervention based on task or physiological pe rformance measures, without mandating precise tracking of a time-dependent traje ctory, poses a significant challenge. This article introduces a new assist-as-ne eded (AAN) robotic training strategy centered on an adaptive velocity field, whi ch guides users smoothly towards a desired path without imposing explicit time c onstraints." Financial supporters for this research include National Council of Humanities, S cience and Technology of Mexico under Grant CONAHCYT. Our news journalists obtained a quote from the research from Center for Research and Advanced Studies, "It promotes participation by reducing assistance based o n task performance and/or muscular effort. Unlike previous works, the low-level controller allows fine-tuning of the robot's accuracy in tracking the velocity f ield and gradually reduces the assistance as the free motion area around the des ired trajectory is approached. This approach facilitates seamless transitions in to and out of the free motion area, where a damping force is provided to ensure stable movements. An additional standout feature is the presence of a move-ahead strategy that avoids shortcuts. Two experiments were conducted to assess the ef fectiveness and advantages of the AAN strategy. Each experiment involved a diffe rent contour-following task with parameters, such as, the shortest distance from the desired path (Experiment 1) and muscular strength (Experiment 2) regulating the level of robotic assistance. In Experiment 1, the proposed strategy was com pared against both a conventional haptic-constraint-based approach and no roboti c assistance. Results indicate that the AAN robotic strategy enables faster task completion, smoother movements, reduced interaction forces, and diminished robo t intervention. Moreover, its real-time adaptability based on task performance a nd physiological data suggests potential benefits for motor learning programs."
SaltilloMexicoNorth and Central Amer icaEmerging TechnologiesMachine LearningRobotRoboticsRobotsCenter fo r Research and Advanced Studies
NETL
NSTL
