Advancements in sensory data analytics and optimization techniques have opened new avenues for enhancing the performance of complex systems across various applications. This thesis explores integrating these methods to improve hierarchical classification and LiDAR-based object tracking, demonstrating their versatility and broad applicability. Hierarchical binary classifiers are a powerful tool for handling multiclass problems, but their effectiveness can diminish as the complexity of the hierarchy increases. Ant Colony Optimization (ACO) is used to optimize the arrangement of classifier blocks within the hierarchy, ensuring the most effective sequence for maximizing detection probability. Adaptive Boosting (AdaBoost) is also employed as a flexible alternative to fixed classifier models, allowing for dynamic adjustment based on the specific characteristics of the data. Monte Carlo simulations conducted on multiclass engine vibration data reveal a significant improvement in detection rates, increasing from \(87.235\%\) to \(90.92\%\) with the optimized structure. In parallel, the thesis investigates the application of sensory data analytics in privacy-sensitive environments using LiDAR technology for object tracking. LiDAR is used to track objects as if a camera were tracking them in the intended closed environment, guaranteeing privacy and confidentiality. Experiments in both static and dynamic scenarios showcase the potential of this technology, particularly when combined with Gaussian Process Regression (GPR) optimized by Particle Swarm Optimization (PSO). This approach achieves precise bounding box coordinate estimation, with a prediction mean square error as low as 0.01. The use of optimization techniques and sensory data across different domains underscores the flexibility and adaptability of these methods.