Chemical engineers are ideally poised to drive the green transition: on a process scale, intensified manufacturing policies that balance sustainability with profits must be determined in real time; on a systems scale, green chemical and energy capacities must be designed and operated to fulfil growing consumer demand; and on a national scale, investment and policy decisions must be formulated to enable an intensified green transition while considering long-term uncertainties in technology development and the economics of chemical manufacturing. Given the increasing amount of digitalization, machine learning (ML) approaches have proved powerful in exploiting data to provide novel insights to challenging scientific problems (e.g., protein folding and weather prediction). Likewise, ML can be used to determine data-driven optimal policies for chemical process operation and development in the green transition. ML is a transformative tool to accelerate the green transition and address the climate crisis. Despite this potential, stakeholders are reluctant to implement ML owing to the black-box problem, whereby the motivation behind the decisions made by these models is opaque to their users. A variety of methods have been introduced to explain the outputs from ML models; this is known as interpretability. By employing interpretable ML-based decision-making tools for developing and operating sustainable chemical processes, they may be deployed quickly, safely, and with lasting process insights. Our research provides a suite of approaches to comprehensively optimize intensified sustainable manufacturing across spatiotemporal scales. Specifically, we want to answer the following questions:1) Given a partial understanding of the mathematical models that govern new and untested sustainable and intensified chemical processes, can their underlying phenomena be learned from data to determine intensified optimal operating policies? 2) What non-market factors can be used to predict the markets involved in sustainable chemical and energy systems? Can these predictions be efficiently embedded into system models to achieve intensified optimal scheduling (i.e., dynamic intensification)? 3) What are the key decisions that influence the uptake of new (sustainable and intensified) chemical projects over time? Can heuristics be learned from data to inform future technological investment and policy?