Chenghao Xiang / 项成昊
把感知，
变成行动。From sensing.
To motion.
我关注机器人与嵌入式系统交汇的地方：连接传感器、边缘计算和运动控制，把工程想法落实为可调试的系统。I work where robotics meets embedded systems—connecting sensors, edge computing and motion control to turn engineering ideas into debuggable systems.
从具体问题出发。Built around real problems.
五个项目，贯穿感知、控制与机电系统。Five projects across perception, control and mechatronic systems.
从深度感知到机器人控制Connecting depth perception to robot control
围绕 RGB-D 感知、边缘视觉与控制接口，探索可调试、可追溯的机器人感知链路。Exploring a debuggable, traceable sensing pipeline that connects RGB-D perception, edge vision and robot control.
在传感器与赛道之间闭环Closing the loop between sensors and track
面向电磁巡线智能车，围绕模拟前端、嵌入式采样和运动控制进行系统协作与调试。System integration and debugging across analog sensing, embedded sampling and motion control for an electromagnetic line-following car.
让感知与电机控制协同工作Bringing sensing and motor control together
以 MSPM0G3507 为核心，将灰度感知、惯性信息和电机驱动连接成自主行驶控制系统。An MSPM0G3507-based control system connecting grayscale sensing, inertial information and motor drivers.
把手势理解带到边缘设备Bringing gesture recognition to the edge
探索基于惯性传感的轻量手势识别，以及从嵌入式推理到相机控制的人机交互链路。Exploring lightweight inertial gesture recognition and an interaction pipeline from embedded inference to camera control.
连接机器人计算与底层运动Connecting robot computing to motion
在自主配送机器人团队项目中，关注上位机与控制器通信、传感集成和运动控制。A collaborative autonomous delivery robot project focused on computer–controller communication, sensing and motion control.
系统的价值， 在接口之间。Good systems live between interfaces.
我的兴趣不止于一个模型或一块控制板，而是理解它们如何在真实约束下共同工作。My interest goes beyond a model or a control board. I want to understand how the pieces work together under real constraints.
了解更多More about me ↗先看系统，再做模块Start with the system
梳理传感、通信、计算与执行之间的职责和约束。Map responsibilities and constraints across sensing, communication, computing and actuation.
让数据与控制可追溯Make behavior traceable
以配置、接口与测试记录支持诊断和工程迭代。Use configurations, interfaces and test records to support diagnosis and iteration.
把验证边界讲清楚Be precise about validation
区分设计、实验与实机结果，用证据支持每一个结论。Distinguish designs, experiments and hardware results. Let evidence support every claim.
把问题继续往前推。Keep the questions open.
从 TinyML 到 VLA，从惯性手势到生物信号。研究条目以探索状态呈现，持续补充方法与验证。From TinyML to VLA, from inertial gestures to biosignals. Research notes describe exploratory work, with methods and validation added as they develop.