The First Differentiable Transfer-Based Algorithm for Discrete MicroLED Repair

TL;DR

This paper introduces the first differentiable transfer-based algorithm for discrete microLED repair, enabling faster, more flexible, and scalable optimization of transfer sequences in high-throughput microLED fabrication.

Contribution

It presents a novel trainable, gradient-based repair algorithm that models discrete platform shifts, outperforming existing local search and RL methods in speed and flexibility.

Findings

50% reduction in transfer steps

Sub-2-minute planning time for large arrays

Superior repair performance compared to local algorithms

Abstract

Laser-enabled selective transfer, a key process in high-throughput microLED fabrication, requires computational models that can plan shift sequences to minimize motion of XY stages and adapt to varying optimization objectives across the substrate. We propose the first repair algorithm based on a differentiable transfer module designed to model discrete shifts of transfer platforms, while remaining trainable via gradient-based optimization. Compared to local proximity searching algorithms, our approach achieves superior repair performance and enables more flexible objective designs, such as minimizing the number of steps. Unlike reinforcement learning (RL)-based approaches, our method eliminates the need for handcrafted feature extractors and trains significantly faster, allowing scalability to large arrays. Experiments show a 50% reduction in transfer steps and sub-2-minute planning time on 2000x2000 arrays. This method provides a practical and adaptable solution for accelerating microLED repair in AR/VR and next-generation display fabrication.