系统基于 Intel 异构计算架构与多路传感器，在边缘侧构建覆盖安全防护、工艺指导与质量评价的一体化智能辅助终端。100% 计算在本地完成。负责硬件系统设计、系统层开发、OS/驱动及硬件加速配置，并完成主程序与视觉及 LLM 子程序的协同调度。

Built an edge-computing intelligent teaching assistant terminal on Intel heterogeneous architecture integrating multi-sensory safety protection, process guidance, and quality evaluation. 100% local computation. Led hardware design, system architecture, OS/driver setup, hardware acceleration, and Vision/LLM orchestration.

设计了一套能够自动识别并复现未知模拟电路幅频特性的测试装置。负责信号处理流程设计与实现、单片机软件开发，构建了多路切换、自校准与自动打表系统，将信号算法高效部署于资源受限 MCU。

Designed an automated testing device capable of identifying and reproducing amplitude-frequency characteristics of unknown analog circuits. Implemented multi-channel switching, self-calibration, and MCU algorithm deployment.

提出一种低成本、非视觉、可交互的智能物体识别与定位方案，利用 RFID 标签与读写器结合信号处理算法实现物体实时状态感知。负责空间信号处理算法设计与实现，开发了基于多天线阵列的 RFID 标签定位算法。

Proposed a low-cost, non-visual interactive object identification and localization scheme combining RFID hardware with signal processing. Developed spatial signal algorithms and multi-antenna array RFID tag positioning.