應用同步定位與建圖技術於工業機器手臂之性能評估

Abstract

性能評估對於確保工業機械手臂的準確性、效率與可靠性具有關鍵意義。傳統量測方法，包括接觸式技術（例如座標量測機及球桿系統）與非接觸式系統（例如雷射追蹤儀及光學座標量測機），雖具備高精度優勢，然而普遍存在成本高昂、安裝複雜及易受環境因素限制等問題。 為優化上述問題，本研究提出一套基於同步定位與建圖技術（SLAM）的性能評估框架，分別採用兩種感測器與演算法：深度相機結合 ORB-SLAM3，與2D LiDAR結合 Cartographer SLAM。兩種系統皆可在無需外部校正基準的情況下實現機械手臂運動軌跡追蹤，提供具成本效益與高度彈性的替代方案。並透過絕對姿態誤差（APE）指標，分析兩種系統在軌跡估計精度與穩定性上的表現，以驗證基於 SLAM 的性能評估方法之可行性。實驗結果顯示，LiDAR-SLAM 系統於軌跡估計中達成 0.0353 mm 的均方根誤差（RMSE），而 VSLAM 系統則展現出在佈署便利性與環境適應性上的輔助優勢。 良好的精度驗證結果顯示，基於 SLAM 的方法具備作為機械手臂性能評估可靠替代方案的潛力。然而，由於 SLAM 系統僅能提供相對於相機座標系下的軌跡資訊，本研究相應地對 ISO 9283 標準中的性能測試定義與計算流程進行了適度調整。上述修正為 SLAM 技術融入標準化性能評估框架奠定基礎。

Performance evaluation is critical for ensuring the accuracy, efficiency, and reliability of industrial robotic arms. Traditional measurement methods offer high precision but are often costly, complex to install, and constrained by environmental factors. To address these limitations, this study proposes a SLAM-based performance evaluation framework, utilizing two sensor configurations: a 2D LiDAR with Cartographer SLAM and a depth camera with ORB-SLAM3. Both systems enable robotic motion tracking without requiring external calibration references, offering a cost-effective and flexible alternative to conventional metrology techniques. To verify the feasibility of SLAM-based performance evaluation, this study analyzes the trajectory estimation accuracy and stability of both systems using the Absolute Pose Error (APE) metric. Experimental results show that the LiDAR-SLAM system achieves a trajectory estimation RMSE of 0.0353 mm, while the VSLAM system provides complementary advantages in ease of deployment and environmental adaptability. The promising accuracy results validate the potential of SLAM-based methods as reliable alternatives for robotic arm performance evaluation. However, since SLAM systems can only provide trajectories relative to the camera coordinate frame, the performance test definitions and computation procedures specified in ISO 9283 were slightly adjusted accordingly. These modifications lay the groundwork for integrating SLAM-based approaches into standardized performance evaluation frameworks.