Intelligent CAD tool for cooperative design of mechanical systems

Smart overview of the Invention

A computer-implemented method is designed to assist design engineers in creating 3D designs of physical objects. The process involves a training phase where sequences of computer-aided design (CAD) operations and their associated metadata are collected. These sequences are converted into tokens, and the metadata into vectors, which are used to train a model stored in a database. In the application phase, the system uses this model to predict and suggest CAD operations based on user input, design constraints, and current operation sequences, providing feedback to the user.

Technical Field

The method falls within the domain of CAD and cooperative assistance systems, specifically targeting the development of virtual 3D object representations in a CAD software environment. It aims to streamline the design process, reduce time and costs, and ease the workload on engineers by offering predictive design assistance.

Technical Background

CAD systems are essential tools for engineers in the product development process, allowing for the creation of virtual models of physical objects. These tools help manipulate the geometric properties of designs to meet functional requirements, such as mechanical load capacities and manufacturing feasibility. The current design process is often manual, time-consuming, and costly, particularly for complex systems requiring coordination across multiple technical domains.

Method Implementation

The method enhances the design process by allowing engineers to add annotations and modify designs to meet complex constraints during the design phase. It is particularly beneficial for both inexperienced and seasoned engineers, as well as those involved in multidisciplinary projects. The system predicts design operations and suggests modifications, helping engineers navigate the intricate requirements of modern product development.

Design Representation and Prediction

The method represents designs as sequences of CAD operations, which are encoded into numerical tokens. Engineers can input additional information, such as manufacturing and cost constraints, as metadata. The system uses this information to predict and suggest design steps to meet specified criteria, facilitating the development of optimal designs. It also allows users to compare their designs with similar ones in the database, offering further insights and design alternatives.