A significant gap exists in our ability to quantitatively predict composite material performance, especially failure processes, due to insufficient techniques to characterize the initiation and propagation of damage spatially and temporally at the 1–100-micron scale.  This topic focuses on pioneering new concepts using emerging innovations in automation, machine learning, sensor, uncertainty, and imaging technologies to address this challenge and produce gold-standard data sets of damage for verification of theoretical concepts and validation of digital engineering models based on a digital twin of the composite (i.e. processing-structure-performance relationship).  The synergy of data from various microstructure and structural characterization methods must isolate the multi-scale mechanistic dependencies that affects failure phenomena with sufficient resolution and precision to provide an understanding to the developing complex stress, environmental, and fracture states.  Damage initiation is a rare event, and thus quantifying and capturing instabilities known to be sudden, unpredictable, catastrophic, and exhibit complex modalities, is critical.  Understanding the impact of multi-scale heterogeneities in microstructure, and developing methods to propagate such uncertainty, are crucial for robust material design and manufacturing tolerance.  In parallel, these microscale investigations should illuminate modeling assumptions that could lead to incorrect predictions, operant mechanisms, or biases in the case of AI/ML.   Example areas of interest (not inclusive) are:  acceleration of testing methods where current approaches can take six months from specimen fabrication to test, pre/post characterization, post-processing of the data, and interpretation;  improving spatial and temporal resolution of X-ray/SEM characterization, and subsequent data-processing, to quantify damage initiation and propagation (both in situ and ex situ);  instantiate workflows and methodologies where previous experiments inform subsequent experiments to drive testing and data efficiencies; develop data plans that enable construction of mechanical property databases based on data from various techniques; and demonstrate inverse engineering approaches that provide Prado Fronts to quantify trade space between microstructure, complex stress state, and damage mechanisms; etc.. In summary, AFRL is seeking innovative experts that can push the boundaries of the material, characterization methods, digitization of experiments, and rapid fabrication processes for highly heterogeneous materials that exhibit anisotropy at the micro to meso scales length scale.  Closing this fundamental gap will provide performance metrics at the subcomponent and component level to tailor multifunctional composite design necessary for future platforms operating through extreme conditions.