Designing of Assembly Framework using Integration of Artificial Intelligence in Assembly Process Planning

The automotive sector is undergoing a major transformation, where traditional, manually assembly planning techniques are about to be integrated with LLM models to reduce work complexity and shorten development cycles. Conventionally, A new carline planning process takes up to 3 years, which involves a large manual referencing and checking of the assembly process with the historical carline, review different parts in a 3D viewer, defining process time analysis, and process documentation in different planning software. This article presents a novel technique using an AI-assisted assembly planning framework that automatically reuses historical carline data from existing carlines and automates assembly planning activities with the use of AI-based intelligent comparison, 3D parts comparison analysis, and real-time autonomous documentation. The framework may help to reduce the initial assembly planning phase from some years to some hours based on an AI processing model. This AI processing model automates product study, DFA checks, PIM (Proactive Improvement Measure) evaluations, Assembly Process Steps (APS) creation, and time estimations. Once AI conducts the study, it will reflect in a dedicated planner dashboard where each point is flagged as yellow/red/green and shows open topics from shopfloor tasks. This allows planners to work only on high-impact points and change assessments only, which leads to a review window for the work. The concept of the reuse-first strategy improves the assembly planning timeline and reduces manual overhead. This study presents a novel approach for the real-time AI integration into 3D viewers and planning tools, for reviewing the workflows and getting continuous feedback on changes or add on very new parts. Future directions include Gen-AI integration for adaptive planning in the automotive sector. Keywords: Artificial Intelligence, Assembly Planning, Change Management, Smart Manufacturing, Digital Twins, Predictive Analytics