Responsibilities Algorithm Prototyping &Validation •Utilize vision models (e.g., YOLO) for real-time detection of parts, tools, hand gestures, and assembly states within a scene. •Explore and fine-tune lightweight LLMs (or Vision-Language Models) to understand assembly process documentation and perform reasoning/decision-making based on visual inputs. Core Application Implementation •Implement intelligent error-proof guidance: the system will provide real-time prompts (via natural language or an AR interface) upon detecting potential errors like incorrect part selection, missing parts, or wrong assembly sequence. •Develop an automated compliance review module for assembly: enabling the system to judge complex logic (e.g., "Have all specified screws been tightened with the correct tool?") based on visual sequences. Requirements •Current Master's or Ph.D. student in Computer Science, Automation, Robotics, or a closely related field. •Proficient in Python, with hands-on project experience using libraries/frameworks like PyTorch/TensorFlow and OpenCV. •Possess a basic understanding and strong interest in Computer Vision (object detection, pose estimation) and Natural Language Processing/Multimodal models. •Preferred: Experience in robotics operation, Augmented Reality (AR), or process planning projects; familiarity with LLM application techniques such as LangChain and Prompt Engineering.