While halide perovskites attract significant academic attention, examples of industrial production at scale are still sparse. In this perspective, we review practical challenges hindering the commercialization of halide perovskites and discuss how machine-learning (ML) tools could help: (1) active-learning algorithms that blend institutional knowledge and human expertise could help stabilize and rapidly update baseline manufacturing processes, (2) computer-imaging methods with ML-based classification tools could help narrow the performance gap between large- and small-area devices, and (3) inference methods could help accelerate root-cause analysis by reconciling multiple data streams and simulations, focusing research efforts on the highest-probability areas. We conclude that to tackle many of these challenges, incremental—not radical—adaptations of existing ML methods are needed. We propose how industry-academic partnerships could help adapt “ready-now” ML tools to specific industry needs, further improve process control by revealing underlying mechanisms, and develop “gamechanger” discovery-oriented algorithms to better navigate the vast spaces of materials choices.

虽然卤化物钙钛矿引起了学术界的广泛关注，但大规模工业生产的例子仍然很少。从这个角度来看，我们回顾了阻碍卤化物钙钛矿商业化的实际挑战，并讨论了机器学习 (ML) 工具如何提供帮助：(1) 融合机构知识和人类专业知识的主动学习算法有助于稳定和快速更新基线制造流程(2) 使用基于 ML 的分类工具的计算机成像方法可以帮助缩小大面积和小面积设备之间的性能差距，(3) 推理方法可以通过协调多个数据流和模拟来帮助加速根本原因分析，将研究工作集中在概率最高的领域。我们的结论是，为了应对其中的许多挑战，需要对现有 ML 方法进行渐进式而非激进式的调整。我们提出了产学合作如何帮助调整“现成的”机器学习工具以适应特定的行业需求，通过揭示底层机制进一步改善过程控制，并开发“游戏改变者”面向发现的算法，以更好地驾驭广阔的材料选择空间。