Scholarlink

<p>This paper presents a flexible, semantic-based inverse kinematics (IK) framework that enables adaptive industrial robot programming across diverse robot types without requiring specialized robotics expertise. Our ontology-enhanced approach integrates semantic knowledge with numerical inverse kinematics methods, allowing task specification at the object level rather than through raw coordinates. Three industrial platforms are addressed: 6-DOF serial manipulators for general pick-and-place operations, 4-DOF SCARA robots for rapid assembly tasks, and 7-DOF redundant manipulators for collaborative obstacle-avoidance tasks. Semantic task specifications are mapped to IK objectives through task-weighted error vectors, a semantically augmented Jacobian, and ontology-driven redundancy resolution. Validation using a 6-DOF manipulator in Unity 3D simulation and physical experiments demonstrated successful automatic adaptation to varying positions without reprogramming, achieving smooth trajectories with 25–30&#xa0;cm path lengths and sub-second completion times (<i>T</i><sub><i>total</i></sub> &lt; 1<i>&#xa0;s</i>) with 60–80 ontological entities. Multi-platform validation achieved &gt; 94% task success rates across four kinematic configurations: physical experiments on the UR5e 6-DOF manipulator and high-fidelity simulation using manufacturer-specified DH parameters for the SCARA YK400XG, ABB IRB 6700, and KUKA LBR iiwa 7-DOF platforms; physical validation on the additional platforms is planned as future work. The system operates effectively in constrained environments through multi-level decision-making and recognizes operational limitations with transparent user feedback. A single-site case study at a partner electronics assembly SME demonstrated 86% setup time reduction (from 3.2&#xa0;weeks to 4.5&#xa0;days, <i>n</i> = 4 new product introductions, ± 0.8&#xa0;days SD), 90% faster product changeover (from 4.2&#xa0;h to 25&#xa0;min, <i>n</i> = 24 changeovers, ± 8&#xa0;min SD), and 99.2% system uptime over twelve weeks (1,440 production hours); these results reflect the specific deployment context and should not be generalized beyond structurally similar small-batch electronics assembly environments without further multi-site validation.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Semantic inverse kinematics for adaptive industrial robot programming

  • 期刊论文

  • 发表日期：

    2026年5月25日

查看全文
  • Japheth Ka’pesha Odira,
  • Jean Bosco Byiringiro

摘要

This paper presents a flexible, semantic-based inverse kinematics (IK) framework that enables adaptive industrial robot programming across diverse robot types without requiring specialized robotics expertise. Our ontology-enhanced approach integrates semantic knowledge with numerical inverse kinematics methods, allowing task specification at the object level rather than through raw coordinates. Three industrial platforms are addressed: 6-DOF serial manipulators for general pick-and-place operations, 4-DOF SCARA robots for rapid assembly tasks, and 7-DOF redundant manipulators for collaborative obstacle-avoidance tasks. Semantic task specifications are mapped to IK objectives through task-weighted error vectors, a semantically augmented Jacobian, and ontology-driven redundancy resolution. Validation using a 6-DOF manipulator in Unity 3D simulation and physical experiments demonstrated successful automatic adaptation to varying positions without reprogramming, achieving smooth trajectories with 25–30 cm path lengths and sub-second completion times (Ttotal < 1 s) with 60–80 ontological entities. Multi-platform validation achieved > 94% task success rates across four kinematic configurations: physical experiments on the UR5e 6-DOF manipulator and high-fidelity simulation using manufacturer-specified DH parameters for the SCARA YK400XG, ABB IRB 6700, and KUKA LBR iiwa 7-DOF platforms; physical validation on the additional platforms is planned as future work. The system operates effectively in constrained environments through multi-level decision-making and recognizes operational limitations with transparent user feedback. A single-site case study at a partner electronics assembly SME demonstrated 86% setup time reduction (from 3.2 weeks to 4.5 days, n = 4 new product introductions, ± 0.8 days SD), 90% faster product changeover (from 4.2 h to 25 min, n = 24 changeovers, ± 8 min SD), and 99.2% system uptime over twelve weeks (1,440 production hours); these results reflect the specific deployment context and should not be generalized beyond structurally similar small-batch electronics assembly environments without further multi-site validation.