A process engineering framework for parameter space optimization in speed breeding

A process engineering framework for parameter space optimization in speed breedingDear Editor, The growing urgency to accelerate crop improvement has renewed interest in trait plasticity-the capacity of plants to adjust growth and development in response to environmental cues (Bradshaw, 2006).This principle underlies speed breeding, in which photoperiod, temperature, light intensity, and CO 2 concentration are manipulated to shorten generation cycles and accelerate crop improvement (Watson et al., 2018).Despite its increasing adoption, optimization of environmental conditions in speed breeding remains largely empirical, relying on trial-and-error experimentation (Ghosh et al., 2018;Gangashetty et al., 2024).Such approaches scale poorly as system complexity increases (Cao et al., 2018).For example, testing six environmental parameters at five levels each yields 15 625 possible combinations, rendering exhaustive experimentation impractical and severely limiting the identification of nonlinear responses and higher-order interactions.Here, we propose a process engineering framework that integrates structured experimental design, machine learning, and explainable artificial intelligence to systematically explore and optimize high-dimensional environmental parameter spaces in speed breeding (Figure 1A).