This paper provides a comprehensive examination of the software ecosystem underpinning modern robotic systems control. As robotics transitions from mechanically dominated platforms to software-centric architectures, understanding the multi-layered software stack becomes critical for engineers, researchers, and industry practitioners. We analyze the global robotics software market, present quantitative performance benchmarks across software layers, and compare leading middleware frameworks including ROS 2, YARP, and proprietary stacks. Our findings demonstrate that software now constitutes 42–45% of total robotic system costs in 2024, and that AI-driven software components have grown from 8.1% to over 30% of software expenditure since 2018. We further document sector-specific adoption patterns across seven industries and identify key challenges in real-time performance, safety certification, and human-robot interaction. The paper concludes with a forward-looking analysis of emerging paradigms including neuromorphic computing, federated learning, and quantum-assisted path planning.