Real industrial decisions (how to schedule production, allocate power and resources, route logistics, or design a network) are high-dimensional, nonconvex problems where gradient-based methods stall or settle for poor local solutions. I design and apply global, derivative-free optimization (genetic algorithms, particle-swarm optimization, and related metaheuristics) to find strong solutions to exactly these problems. This pillar is grounded in years of deployed industry engineering, much of it proprietary rather than published: optimization software for production planning, power consumption, transportation, and building/HVAC systems, including a genetic algorithm for chemical-industry production scheduling. Published examples include heuristic network design for content caching in 5G and a new encoding method for complex production systems.
As co-founder and R&D engineer at GAAS (Genetic Algorithms Application Space) in Jordan (2010–2015), I engineered global-optimization algorithms for real industrial operations: HVAC control, building management systems, production planning, and scheduling. See also Industry Experience.