Robotics researchers have introduced an adaptive optimization method designed to resolve inverse kinematics challenges for redundant inspection robots operating on aircraft wings. The technique pairs an adaptive search-space module with the second-generation nondominated sorting genetic algorithm, a framework designated by the authors as ASM-NSGA-II. In computational evaluations across four distinct trajectory paths, the algorithm demonstrated substantial reductions in mean-fitness metrics compared to standard baseline solvers while keeping solution runtimes comparable to existing numerical approaches.
Automated wing scanning requires robotic arms to maintain continuous joint configurations, ensure feasible probe orientations, and avoid mechanical limits across extended multi-axis movements. To address these demands, investigators formulated a specialized kinematic pipeline for a ten-degree-of-freedom robotic setup consisting of a four-axis external mobile base coupled with a six-axis industrial articulated arm.
To manage this dimensional complexity without excessive computational burden, the study decomposed the robot’s forward-kinematic chain. Rather than solving for all ten degrees of freedom simultaneously through generic global search routines, the algorithm optimizes only the four redundant parameters associated with the external axes. Analytical inverse kinematics for the six-axis industrial arm then reconstructs candidate joint configurations, which are systematically screened against strict physical joint limits and probe-pose constraints.
The multiobjective selection routine within the solver evaluates candidate configurations using three guiding criteria: joint continuity, joint-limit margin, and mechanical manipulability. The adaptive search-space module enhances solver efficiency by scoring nested candidate regions through repeated presampling stages. Furthermore, normalized similarity between adjacent poses dynamically controls the migration of solutions from the preceding population, concentrating computational resources within the most promising zones.
During algorithmic execution, the adaptive search-space module dictates the designated sampling boundaries, while the core NSGA-II mechanism filters and preserves nondominated Pareto-optimal solutions within that confined area. In benchmark tests comprising ten independent experimental runs on each of four simulated paths, the hybrid approach achieved mean-fitness reductions ranging from 41.6 to 51.5 percent relative to the standard NSGA-II framework. When contrasted with the published T-IK method, the routine produced mean-fitness decreases between 6.2 and 25.0 percent.
The computational cost of achieving these refined joint trajectories remained balanced against existing standards. The mean solution time of the ASM-NSGA-II method differed from the established T-IK approach by no more than 8.7 percent across the evaluation paths.
The research team connected mesh-based probe pose construction directly to joint-sequence execution within an integrated MATLAB and CoppeliaSim simulation environment. This virtual demonstration simulated an aircraft wing-surface scanning task, providing model-level verification of continuous motion profiles.
The study authors noted that while computer simulations demonstrated improved aggregate kinematic trade-offs for the tested configurations and paths, physical tracking precision, real-time closed-loop operation, and comprehensive collision safety remain unverified and require future experimental validation on physical robotic hardware.
Conducted by researchers at the University of Electronic Science and Technology of China, the study has been released as a preprint under review for the journal Mechanical Sciences.
Subject of Research: Inverse kinematics optimization for redundant aircraft wing inspection robotics
Article Title: Inverse Kinematics for a Redundant Wing-Inspection Robot via Multiobjective Optimization with an Improved Second-Generation Nondominated Sorting Genetic Algorithm
Article References: Dai, X., Zhou, X., Tu, N., He, Y., Zhang, Y., & Zhao, P. (2026). Inverse Kinematics for a Redundant Wing-Inspection Robot via Multiobjective Optimization with an Improved Second-Generation Nondominated Sorting Genetic Algorithm. https://doi.org/10.5194/ms-2026-198
Image Credits: AI Generated
DOI: 10.5194/ms-2026-198
Keywords: inverse kinematics, wing-inspection robot, multiobjective optimization, NSGA-II, adaptive search space, kinematic redundancy, aircraft maintenance, manipulability, joint limits, trajectory planning, degrees of freedom, MATLAB, CoppeliaSim, robotic inspection, motion planning
News Source: Denise Maddox. (November 19, 2026). Adaptive Genetic Algorithm Refines Motion Planning for Redundant Wing-Inspection Robots. Scienmag.



