Physical algorithmic and reservoir computing: Foundations for mechanically intelligent structures

We are witnessing a rapid emergence of mechanically intelligent structures and material systems capable of information processing, learning, decision-making, and actuation without centralized digital computing and learning. Central to this development is the creation of physical substrates that can compute—encoding, processing, and decoding information through their intrinsic physics. Among the many physical computing approaches proposed to date, Physical Algorithmic Computing and Physical Reservoir Computing have emerged as two of the most promising frameworks because of their well-defined computational architectures, programmability, and high computational capacity. This special issue, “Embodied Physical Computing and Mechanical Intelligence,” highlights recent advances in these rapidly evolving areas through a collection of original research articles spanning responsive materials, adaptive structures, soft robotics, and embodied computing. In this brief Perspective, we discuss the current status of these two physical computing paradigms, identify key challenges and emerging opportunities, and outline future research directions in physical adaptability, modularity, and new performance metrics.