Recent research:
This body work consolidates advances in autonomous routing, intelligent control, and AI-driven optimization across robotics, industrial systems, and power/energy domains. A common thread is the integration of learning or heuristic optimization with practical control implementations to enhance efficiency, adaptability, and reliability in real-world settings. Route optimization for autonomous ground vehicles emphasizes real-time planning under spatiotemporal and curvature constraints, leveraging PRISMA-based literature review and metaheuristics (e.g., Humpback Whale Optimization, Firefly, PSO) to identify trends and gaps for flexible navigation in agriculture, logistics, and surveillance. In multivariable industrial contexts, the work demonstrates decoupled, tuned control via RGA in Siemens TIA environments, validated against Matlab-Simulink, improving efficiency and stability. UAV swarms and vision-based trajectory systems address practical deployment hurdles, including collision avoidance with Deep Learning and ArUco markers, supported by controlled experiments to quantify flight time and evasion. Deep learning and neuro-fuzzy approaches appear across motor control and process automation: adaptive PI and FOPID controllers optimized with bio-inspired algorithms (AF, ABC, GA, PSO, IWO) on ARM/STM32 platforms, with performance metrics such as IAE, ITAE, and Wilcoxon tests demonstrating gains in accuracy and energy efficiency. Additional work explores intelligent assistants and AR-enabled predictive maintenance, highlighting reliability and real-time cloud-enabled monitoring; cyber-incident ML classification for ICS security; and vision-based assistive devices for visually impaired users. Collectively, the research advances how AI, fuzzy/neural control, and metaheuristics can be embedded in affordable hardware (ARM/STM32, Raspberry Pi, Arduino) to yield robust, energy-aware, and perceptive control in robotics, manufacturing, energy systems, and safety-critical applications.