About关于
Alvin Li
The whole world is just one giant embedded system这个世界就是一台巨大的嵌入式
I'm Alvin Li. I studied telecommunications engineering and I'm now doing an MSc in Robotics and Intelligent Systems at NTU. Most days I draw schematics, lay out boards and write firmware; by the end of my bachelor's that added up to more than 30 PCBs, up to six layers. In the lab I now build sensing hardware for robots. Anything that moves on its own, from a traffic light to a lithography machine, hides control electronics and embedded code, and I want to understand how all of it works.我是 Alvin Li，本科读通信工程，现在在 NTU 读机器人与智能系统硕士。平时画原理图、布板、写固件，本科四年下来画了 30 多块 PCB，最多做到 6 层，现在在实验室里给机器人做传感硬件。从路口的红绿灯到光刻机，能自己动起来的东西里都藏着电控和嵌入式，它们到底是怎么运转的，一直是最吸引我的问题。
Traffic lights, elevators, washing machines, robot vacuums, drones, EV motor drives, robot arms, satellites and lithography machines are all embedded systems.红绿灯、电梯、洗衣机、扫地机器人、无人机、电车电驱、机械臂、卫星和光刻机，都是嵌入式系统。
The Road So Far我走的路
The turn toward embedded work happened slowly. Telecom foundation courses introduced me to microcontrollers, circuits and motor control, and I got hooked. I kept wanting to know how the world actually runs.转到嵌入式这条路，是一点点发生的。通信工程的专业通识课里学到单片机、电路和电控，越学越着迷，总想知道这个世界到底是怎么运转起来的。
The answer that stuck: the world is one huge embedded system. Every automatic thing around us, from the traffic light on the corner to the most precise lithography machine on Earth, depends on control electronics, embedded code and hardware. After that there was no stopping. I taught myself the basics from online roadmaps, slowly started building projects, and never left.学完才发现，这个世界就是一个巨大的嵌入式系统。身边的红绿灯也好，全世界最精密的光刻机也好，所有能自动控制的东西都离不开电控、嵌入式和硬件。从此一发不可收拾，先照着网上的学习路线自学入门，再慢慢开始做项目，就再也离不开这个方向了。
Alongside the projects, I co-founded the XJTLU Hardware Studio and led its electronics: every circuit-level piece of hardware went through me, while modelling and mechanics sat with others.做项目的同时，也参与创立了 XJTLU Hardware Studio，担任电子硬件主理人。工作室里所有电子电路方向的硬件都归我管，建模和机械交给其他伙伴。
Line tracker and digit counter循迹小车与数码管计数器
Two MEC104 course builds. The first time I soldered, wired and wrote Arduino code until something actually ran.MEC104 课上的两次实训。第一次自己焊板、接线、写 Arduino 代码，一直折腾到东西真的跑起来。
DIY laptop coolerDIY 笔记本散热器
My first project of my own: an ESP32 between the computer and a fan, so the laptop could drive its own cooling.第一个完全由自己起意的项目。用 ESP32 把电脑和风扇连起来，做成一台能被电脑控制的压风式散热器。
STM32 intern, Tianjin Jintie天津津铁 STM32 实习
A month that took me from Arduino-style pin calls down to GPIO, interrupts, timers, ADC, PWM and UART on bare metal.一个月的嵌入式实习，从 Arduino 式的高低电平一路补到 GPIO、中断、定时器、ADC、PWM 和串口。
Qt/C++ intern, Nanjing Turing南京图灵 Qt 开发实习
Three weeks inside the Seamly2D codebase: my first features in a large open-source project, then Windows and macOS packaging.三周泡在 Seamly2D 的代码里，第一次在大型开源项目里加功能，再把它打包成 Windows 和 macOS 的安装包。
Phase-change cooling controllers, Cirro Tech卷云科技相变散热器控制系统
Hardware and firmware for the BaseUnit and the three-board ACUnit, carried from schematic all the way to a running refrigeration loop.独立负责 BaseUnit 和三板 ACUnit 的硬件与固件，从原理图一路做到整条制冷链路跑通。
Custom driver board and sensorless FOC自制驱动板与无感 FOC
My own driver board spins a compressor open-loop. The sensorless closed loop is still on the bench.自己画的驱动板已经把压缩机开环转起来了，无感闭环还在台架上慢慢磨。
Claude Chime power boardClaude Chime 电源板
A commissioned board I took from requirements through schematic, PCB, hand assembly and bring-up on my own.一块委托设计的电源板，从需求拆解、原理图、PCB 到手工焊接和上电调试，都由我一个人完成。
Work in Progress在做的事
I'm at NTU for the MSc in Robotics and Intelligent Systems, working on wearable muscle-sensing hardware in Assoc. Prof. Wang Yifan's Robotic Materials Group. My dissertation sits right there: a smaller, wearable bioimpedance and EMG acquisition platform, which for me means redesigning the mixed-signal front end, building the embedded data acquisition and getting the PCB and firmware to work as one. It's human–robot interaction work: bioimpedance becomes raw input for robot models, so a robot can better understand how a person applies force while gripping or moving.现在在 NTU 读机器人与智能系统硕士，在王一凡副教授的 Robotic Materials Group 里做可穿戴的肌肉传感硬件。毕业论文也落在这里，要做一套更小型化的可穿戴生物电阻抗与肌电（EMG）采集平台，手上的活是重新设计混合信号前端、搭嵌入式数据采集，再让 PCB 和固件真正配合起来。这属于人机交互方向，把电阻抗当作原始输入交给机器人模型，让机器人更懂人在抓握或做动作时到底是怎么发力的。
Two families of algorithms keep showing up in my projects. On the control side there's PID, LQR and, more and more, NMPC. On the signal side there's EMA, lock-in amplification, SINC3 filtering and Kalman filtering with the EKF. The more I use them, the more they feel like two halves of the same job.做项目时常用到两类算法。一类管控制，PID、LQR，再到更主流一点的 NMPC；另一类管信号，EMA、锁相放大、SINC3 滤波，还有卡尔曼滤波和 EKF。用得越多，越觉得这两类东西是相辅相成的。
Further out, I'm drawn to embodied intelligence and world models: how robots come to understand the world. The plan is to get the hardware solid first, then use the data it collects for reinforcement learning and model training. Which is why I basically live in the lab right now (sigh).再往后，想往具身智能和世界模型走，弄明白机器人是怎么理解这个世界的。眼下的打算是先把硬件做扎实，以后再拿采集到的数据去做强化学习和模型训练。所以现在每天都泡在实验室里（悲）。
A student pass caps work at 16 hours a week, so a proper full-time internship is off the table for now. I'm looking for research assistant or part-time roles that fit those hours, partly to keep myself fed (ha). If you have one, I'd love to hear from you.学生签证每周最多只能工作 16 小时，正经的 full-time 实习暂时找不了，所以在找能配合这个工时的 Research Assistant 或兼职机会，顺便补贴一下生活（笑）。如果你那边刚好有合适的位置，欢迎来聊。
Off the Clock认真之余
- Cutting weight减肥中
- Gym健身
- Badminton羽毛球
- Road cycling公路车
- Piano钢琴
- Chess下棋
- Talking with AI跟 AI 聊天
- Writing, now and then偶尔写作
I love chasing new technology and genuinely enjoy building things together with AI; this site is one of them, designed and built by me with AI at my side. AI has raised the floor of everyone's life and lifted the ceiling of the world at the same time.喜欢追前沿技术，也很享受跟 AI 一起把东西做出来的感觉，这个网站就是我和 AI 一起从零搭起来的。对中文互联网上流行的“薄肌黄毛，大水大睡，AI 管理生活”理论，本人深表认同。AI 平等地抬高了每个人人生的下限，也拔高了这个世界的上限。
That is exactly why I won't hand it my whole brain; lean on it long enough and the thinking muscles waste away. AI can do your work and write your code, but how you think and how you face the world are still yours to decide. I'd rather it taught me to fish. Plenty of world-changing things began as one ordinary thought, and that is the part AI probably can't help with. Reading widely and travelling far still matter.也正因为这样，更不想把脑子整个交给它，用久了是真的会退化。AI 可以替你干活、替你写代码，但该怎么思考、怎么面对这个世界，只能自己决定。最好让 AI 授人以渔，别只顾着授人以鱼。很多改变世界的东西，起点往往只是一个普通的念头，这一步 AI 大概帮不上忙。所以读万卷书、行万里路，放在今天依然重要。