Conventional processors face fundamental bottlenecks when deployed in resource‑constrained aerospace systems that demand onboard intelligence. This review examines brain‑inspired neuromorphic computing as an alternative, focusing on digital neuromorphic processors with embedded central processing unit (CPU) control, multi‑core architectures, hybrid compute fabrics, and on‑chip learning capabilities for space applications. This review first translates aerospace constraints into chip‑level design requirements, such as low‑power event‑driven computation, sparse communication, hybrid model support, and incremental on‑chip learning. Based on these requirements, we survey four key technical routes in neuromorphic chip designs: heterogeneous control architectures, network‑on‑chip interconnects, hybrid compute fabrics, and on‑chip learning engines. A comparative evaluation of fabricated chips reveals a clear trend toward larger scale, higher energy efficiency, and greater model flexibility. Validation cases across four aerospace domains are examined, though most remain ground‑based with little flight heritage. This review identifies four persistent challenges and outlines corresponding future research lines, including scale, learning, validation and ecosystem aspects. It concludes that neuromorphic chips offer a viable path toward low‑power, adaptive, and robust onboard intelligence, provided that close collaboration bridges the remaining gaps.

Soft materials are attracting increasing attention in aerospace engineering because they provide combinations of compliance, multifunctionality, and environmental adaptability that are difficult to achieve simultaneously with conventional rigid materials. This review summarizes recent progress in the use of elastomers, hydrogels, aerogels, shape-memory polymers, and related composites for aerospace applications, with emphasis on impact and vibration mitigation, sealing and leakage control, intelligent actuation and deployable structures, thermal insulation and management, radiation protection, flexible sensing, and multifunctional integration. Across these areas, the role of soft materials is shifting from passive auxiliaries toward functional material platforms that couple structure, actuation, sensing, transport regulation, and protection. The literature also shows that several key barriers remain unresolved. Soft materials in aerospace must operate under coupled thermal, mechanical, chemical, vacuum, radiation, and atomic-oxygen environments, yet current understanding of long-term degradation, cross-scale failure, and function retention under such conditions remains limited. In addition, many reported applications are still evaluated mainly through laboratory-scale demonstrations, leaving a substantial gap between proof-of-concept performance and mission-relevant validation. Predictive theories linking molecular design, microstructural evolution, and system-level behavior are still insufficient. This review identifies four closely related directions for future progress: coupled multiphysical and mechano-chemical theories for deformation, aging, and failure; accelerated testing and lifetime prediction methods linked to realistic service conditions; lightweight system architectures integrating structure, actuation, sensing, and protection; and AI-assisted design combined with 4D printing and other advanced manufacturing approaches for rapid optimization and mission-specific customization. Overall, soft materials should be developed not merely as discrete functional elements, but as enabling platforms for future aerospace systems under extreme environments.

A novel physics-informed recurrent neural network (PIRNN) model has been developed for data-driven constitutive modeling of metallic materials under high-strain-rate loading. The proposed PIRNN framework integrates a gated recurrent unit (GRU) network and a fully connected neural network to capture the rate-dependent stress-strain behavior directly from synthetic stress and strain sequences generated from the Johnson-Cook constitutive model. Notably, the inherent memory capability of the GRU model enables implicit representation of arbitrary loading history, without the need for explicitly accounting for internal state variables or prescribed constitutive relations, while the fully connected neural network is utilized to estimate Helmholtz free energy under general loading conditions. Thermodynamic admissibility is ensured by enforcing Helmholtz free energy positivity and the dissipation inequality within the loss function. Validation at the material point level against direct numerical solutions and experimental data demonstrates the accurate predictive capability of the PIRNN under both multiaxial and uniaxial loading conditions. The PIRNN is further implemented in ABAQUS for structural simulations to demonstrate its applicability in structural analyses.