Intelligence-enabled smart inhalable biomaterials for chronic inflammatory airway diseases: integrating responsive carrier engineering, lung-on-chip validation, and patient-specific pulmonary modeling

Chronic inflammatory airway diseases, particularly asthma and chronic obstructive pulmonary disease (COPD), are commonly treated by inhalation, yet the nominal dose placed in an inhaler often differs substantially from the dose that reaches diseased airway regions, crosses mucus, enters target cells, and remains pharmacologically active. Airway narrowing, heterogeneous ventilation, mucus plugging, impaired mucociliary clearance, oxidative and proteolytic stress, epithelial barrier dysfunction, inflammatory-cell diversity, device characteristics, and patient-specific inhalation maneuvers jointly shape therapeutic exposure. Smart inhalable biomaterials can be designed to address some of these barriers by controlling transport, retention, activation, or release, but many systems are still optimized primarily at the nanoparticle level and evaluated in static models that do not capture realistic aerosol performance or patient-specific deposition. This review develops an intelligence-enabled bioengineering framework linking disease-informed material design with machine-learning-assisted formulation optimization, formulation-device-patient coupling, lung-on-chip validation, and patient-specific pulmonary modeling. Within this framework, airway features such as mucus architecture, redox gradients, protease activity, epithelial permeability, immune-cell state, extracellular-matrix remodeling, and regional airflow are considered measurable variables that can inform material design and evaluation. The material platforms considered include mucus-penetrating and mucoadhesive carriers, responsive materials, transformable nano-in-micro systems, polymeric and lipid nanoparticles, biomimetic carriers, extracellular vesicles, and systems for pulmonary delivery of nucleic acids and biologics. Their development is considered alongside structured experimental design, machine learning, active learning, mechanistic aerosol modeling, physics-informed surrogates, multi-objective optimization, and uncertainty analysis. Lung-on-chip systems provide human-relevant settings in which deposited dose, mucus transport, barrier function, immune responses, and repeated-exposure effects can be examined under air-liquid-interface and breathing-like conditions. Patient-specific computational fluid dynamics and reduced-order models can complement these platforms by predicting regional deposition, whereas the term digital twin should be reserved for longitudinally updated, uncertainty-aware models linked to actionable decisions. Together, these approaches support a closed-loop design-delivery-validation strategy in which material performance, aerosol behavior, biological response, and patient-specific deposition are evaluated as connected rather than separate development problems.