Capstone Final Project Report - AutoStack; AI-Driven Automation and the Reorganization of Laboratory Work 6 views

Modern scientific laboratories are increasingly shaped by artificial intelligence and automation. These technologies improve efficiency and precision. However, they also change how researchers do their work. Many laboratory tasks once required long training and careful manual control. Today, automated systems can assist or perform these tasks. For example, materials science researchers must align microscopic samples under a microscope. This process requires steady hands and long practice. At the same time, computer vision and robotics now offer automated solutions. These systems can perform alignment more consistently. This situation raises an important question. How can laboratories use automation while still preserving human expertise and accessibility? This question connects both of my projects. My technical project builds a low-cost automation system. My STS research studies how automation changes human roles and skills. My technical project focuses on a system called AutoStack. This system automates the alignment of two-dimensional materials. The project begins with a clear problem. Researchers must align tiny flakes under a microscope. This task often happens inside glove boxes. These environments limit visibility and movement. As a result, the work is slow and difficult. Success depends on experience. AutoStack addresses this challenge by combining computer vision and motorized stages. The system detects material flakes and suggests positioning. The system also allows user control through a controller interface. During testing, the system performed positioning tasks more consistently than manual methods. The system reduced the need for constant manual control. The system improved repeatability across trials. However, the system does not replace the user. Users still monitor outputs and correct errors. Users also refine alignment when needed. These results show a clear pattern. Automation improves efficiency but still depends on human input. The project also shows that low-cost systems can expand access to advanced tools. My STS research examines how automation reshapes laboratory work. The research asks a central question. How does automation change researchers’ roles and expertise? The research also asks how researchers shape these systems in return. This study uses a sociotechnical framework. This framework focuses on mutual shaping between humans and technology. The findings show that automation does not simply replace labor. Instead, automation reorganizes work. Researchers shift from manual tasks to supervision and decision-making. For example, researchers no longer perform every alignment step by hand. Instead, they monitor system outputs and intervene when needed. This shift also changes expertise. Traditional work requires manual precision and practice. Automated work requires system understanding and interpretation. Researchers must understand how systems operate. Researchers must also detect errors and make adjustments. The study also shows that interaction remains important. Researchers respond to system behavior and adjust parameters. This interaction shapes how systems function in practice. The research also highlights unequal access to automation. Many systems are expensive and require advanced infrastructure. Smaller laboratories often cannot adopt them. This limitation shows the importance of low-cost solutions. Overall, the study shows that automation is a sociotechnical process. Humans and systems continuously shape each other.

BS (Bachelor of Science)

AutoStack; Artificial Intelligence; Automation; 2D Materials; Laboratory

School of Engineering and Applied Science Bachelor of Science in Computer Engineering Technical Advisor: Andreas Beling STS Advisor: Kent Wayland Technical Team Members: Frank Wu, Yingming Ma, Feixiang Liao

English

All rights reserved by the author (no additional license for public reuse)

2026-05-09