Maker · M.Eng. (UCAS) · Ph.D. Student (HKUST)
Ma Weilin is a maker and holds a Master’s degree in Electronic Information from the University of Chinese Academy of Sciences (UCAS). He is currently a PhD student in Physics at the Hong Kong University of Science and Technology (HKUST). His research focuses on the development of low-cost scientific instrumentation. Representative projects include OpenSTM and a micro-profilometer. During his PhD studies, he is working on the prototyping of Superconducting Charge Qubit Microscopy (SCQM) under the supervision of Prof. Berthold Jaeck.
Research interests: Quantum Sensing, Scanning Probe Microscopy, Low-Cost Alternative Scientific Instrumentation
Hong Kong University of Science and Technology
2025 — Present
Ph.D. in Physics
University of Chinese Academy of Sciences
2022 — 2025
M.Eng. in Electronic Information Engineering
Wuyi University
2018 — 2022
B.S. in Electronic Information Engineering
A 300 USD(2000 CNY) cost scanning tunneling microscope(STM).
This project is an open-source, low-cost STM. The system operates at ±15 V(maximum) and integrates a tip-approach algorithm, enabling the STM to establish a tunneling current within one minute. It is particularly suitable for educational use. The pictures below show the HOPG sample scanned by OpenSTM.
A confocal displacement sensor based profilometer.
This project developed a 2D piezoelectric motor with a 100 nm resolution. With an integrated confocal displacement sensor, the system can measure a sample’s surface profile with 38 nm depth resolution and 8 µm XY resolution. The pictures below show the measurement results for a chip die and anodized aluminum oxide.
Resolution-improved, low-cost self-mixing interferometry (SMI)-based displacement sensor for piezoelectric materials.
This project developed an SMI-based displacement sensor for piezoelectric materials. With the proposed fusion algorithm, the RMSE was reduced by 30.1% and the maximum error by 45.5% compared with the original SMI sensor, achieving an accuracy of 120.7 nm and an RMSE of 45 nm. The resolution reached 22 nm, compared with 327.5 nm for the original SMI sensor.
Customized text-to-text chatbot.
The chatbot is an encoder–decoder text-to-text model. The encoder uses a BiLSTM, while the decoder uses an LSTM with an attention mechanism. The bot can learn from group chat history and chat with users after training.
Tip: You can click any image to open it in a new tab.
I write research notes and project retrospectives here: Visit my blog