In conventional manufacturing processes, various types of data are managed separately by each process. These include data created during each process as well as the details and results of decisions that were made. Because of this, it is difficult for engineers working on these processes to access or utilize data of other processes, which impedes the sharing of knowhow among engineers and leads to problems such as rework in downstream processes.
FUJIFILM Business Innovation has been working on research to utilize its 3D printing data format FAV,*1 which can retain not only data designed using 3D CAD but also complex three-dimensional data (e.g., internal structure, color, material, and connection strength) of three-dimensional objects. Discussions with people from a diverse range of fields to promote practical applications of FAV led us to discover a demand for integrated handling of various types of three-dimensional data other than that used for 3D printing. To meet this demand, the specifications of the FAV format have been expanded with the update to version 1.1a. Moreover, with the application of these expanded specifications, a 3D data platform enabling the integrated management of 3D data created in manufacturing processes has been established (Fig. 1).
Because it is expected to be utilized for 3D printers and has been recognized as valuable due to the expansion of its scope of utilization, FAV as a versatile 3D data format is now established as a Japanese Industrial Standard (JIS).*2 In the future, FAV is expected to be utilized more widely in the manufacturing field, contributing to the establishment and dissemination of more efficient manufacturing processes.
Many different types of data generated during manufacturing processes, each with different resolutions and formats (e.g., simulation results (csv) or measurement results (point clouds)), can be registered as attributes of voxels in the FAV format. Although voxel data processing has previously been viewed as having problems due to its large data size and the decline in reproducibility of curved surfaces or distinctive features such as corners, the application of image processing technology developed by FUJIFILM Business Innovation has made the processing time 100 to 1,000 times faster while maintaining the accuracy of shapes. This technology makes it possible for various types of data that have conventionally been managed separately by each process (e.g., pre-manufacturing simulation information, configuration data of manufacturing equipment, and post-manufacturing measurement result data) to be integrated with high accuracy into a single data file, thus enabling the data to be efficiently shared and utilized across all processes.
By creating FAV data for each product and establishing a database of FAV data for all products, it becomes possible for users to search for similar data created in the past. When data such as configuration data of manufacturing equipment and measurement data is managed as a single integrated FAV data file, it can be referred to in advance, making it possible to prevent problems which may occur in the manufacturing process.
Moreover, information which was previously managed separately by each process, such as the intention behind a design, manufacturing restrictions, and tolerances defined by manufactured parts, is made accessible by other processes, resulting in improved productivity through accelerated decision-making and prevention of rework. In addition, with AI utilizing large amounts of FAV data as big data, FAV is expected to serve as a new means of communication, taking engineers’ knowhow, which is often difficult to verbalize, and expressing it in the form of data analysis results.
- *1 FAV (abbreviation for “fabricatable voxel”): expresses an entire three-dimensional object by arranging basic configuration units (voxels), which retain various types of information regarding the object. Version 1.1a of the FAV data format was established as a Japanese Industrial Standard (JIS).
- *2 JIS B9442: Specification of FAV format for 3D models