Industrial Unmanned Aerial Vehicle (UAV) applications requiring centimeter-level precision face fundamental limitations when traditional control methods encounter complex aerodynamic environments and safety constraints. This paper makes three contributions: (1) a multi-component architecture achieving 38.6 percentage point improvement over standard reinforcement learning for UAV precision control; (2) adaptive three-phase safety constraint optimization with formal convergence guarantees enabling 11.6 percentage point precision gain; (3) analysis of component interactions showing that decision transformers drive performance gains while physics constraints enforce realism. The framework integrates Physics-Informed Neural Networks (PINNs), Neural Operators, Decision Transformers, and Control Barrier Functions (CBFs). Through ablation studies across eight configurations, we evaluate how different physics-informed components interact. The framework achieves 91.9% precision within 10 cm tolerance, 55.3% within 5 cm, and 5.5% within 2 cm while maintaining 99.9% physics consistency. The complete system requires 5.6 h of training on consumer hardware, making practical deployment feasible. Our analysis shows that carefully calibrated safety-precision optimization, rather than rigid constraint enforcement, produces better performance while maintaining safety guarantees in precision-critical autonomous systems.

Testing self-driving cars (SDCs) requires extensive simulation-based testing, making efficient test case selection important. This paper presents two approaches for test case selection in SDC testing: a curvature-based selector that analyzes road geometry and a graph neural network (GNN) based selector that learns failure patterns. The curvature-based approach uses road geometry analysis, turn detection, and group-based selection strategies, while the GNN approach has a four-layer neural architecture with feature engineering for predicting test failures. Both approaches are implemented and evaluated as part of the ICST 2025 Tool Competition for SDC testing. Our experimental results show that the GNN selector achieves superior computational efficiency (initialization: 1.45s vs 15.27s, selection: 0.30s vs 4.06s) and a better time-to-fault ratio (209.78 vs 239.48), while the curvature-based selector demonstrates stronger fault detection capabilities with a higher fault-to-selection ratio (0.214 vs 0.177). Both approaches maintain comparable diversity scores (0.037 and 0.039 respectively), demonstrating their effectiveness in achieving comprehensive test coverage. The comparative analysis provides insights into the strengths of geometric analysis and machine learning approaches in SDC test selection.