SignalPro and Macnica demo AI-powered golf robot arm
SignalPro Technology and Macnica held an exhibition on November 18 featuring an Epson robot arm that combined millimeter-wave radar and stereo cameras to play golf. The system used NVIDIA Jetson Orin Nano 8GB with CUDA and TensorRT for edge processing, while MSI EdgeXpert trained sensor models and NVIDIA Nemotron estimated shot decisions. SignalPro said the system interpreted randomly generated targets and scores, adjusted swing force and speed, and hit the designated targets. Its claims about sensing accuracy, inference speed, and stability were presented as the company's account.
The demonstration showed a Physical AI workflow combining multimodal sensors, edge computing, and an AI model to make and execute golf-shot decisions. The report attributes the system's performance and stability claims to SignalPro.


希格諾科技股份有限公司(SignalPro Technology)於 11 月 18 日與茂綸股份有限公司共同舉辦展覽活動,展示 Physical AI 於實體場域的應用成果。此次展出的核心亮點,是在機器手臂上整合毫米波雷達與雙目相機,讓機器手臂具備感測與判讀能力,實際執行揮擊打高爾夫球的動作。
活動過程中,每一局擊球的目標位置與分數皆為隨機產生,Epson 機器手臂搭載的 AI 系統需即時判讀所需的擊打力度與速度,並據此調整揮桿動作,將球精準打向系統設定的目標位置。此設計目的在於呈現 Physical AI 如何讓機器具備「感知環境、即時決策、精準執行」的完整能力。
#SignalPro Technology#Macnica#Epson#NVIDIA#NVIDIA Jetson Orin Nano 8GB#NVIDIA CUDA#NVIDIA TensorRT#MSI EdgeXpert#NVIDIA GB10 Grace Blackwell#NVIDIA Nemotron#TensorRT-LLM#NVFP4