Innovation in Motion: Joining Hands with Keio University to Create FAV, a World-First 3D Data Format

Potential Becomes Reality: 3D Printing in a New Era of Manufacturing Masahiko Fujii, Tomonari Takahashi Marking Technology Laboratory, Research & Technology Group

In April 2015, Fuji Xerox entered into a joint research effort to investigate the handling of 3D data, pairing up with the Social Fabrication Lab of Keio Research Institute at Shonan-Fujisawa Campus (SFC) (headed by Professor Hiroya Tanaka of Keio University). Within less than a year, the combination of their efforts resulted in the birth of FAVNote 1, the world’s firstNote 2 data format for 3D printing capable of retaining complex information. Soon after, in July 2016, the specifications of this new format were released online. In this volume, we talk to the members involved about the collaboration and the key factors that allowed them to achieve their goal in such a short span of time.

Pursuit of the ideal 3D printing data format

Marking Technology Laboratory, Research & Technology Group
Masahiko Fujii (left) Tomonari Takahashi (right)

Ten years ago, 3D printers typically printed objects using the stereolithography process. Their most common application was “rapid prototyping,” meaning they were primarily used to create mock-ups and visual models. This, however, meant that 3D printers featured only briefly in the overall process of development and manufacturing. Since then, technological developments have made 3D printers able to process a wider range of material information, and consequently the scope of their application in development and manufacturing has widened considerably.

Even in the low-end market, consumers are becoming increasingly accustomed to 3D printers as a day-to-day manufacturing tool, and it is expected that further growth in this market will be seen in years to come. At Fuji Xerox, an investigation was carried out into the ways that we could contribute to these developing 3D printing markets, and in this investigation, the following major issues were identified:

  • The most common data formats for 3D printing use polygons (triangles) and describe only the surface geometry of 3D objects. Information necessary for full-color, multi-material objects and complex internal structures is not specified within these polygons, despite technological advancements that have made such expression technically feasible.
  • The data-flow steps (such as complex intermediate processing) that are currently necessary in outputting data to 3D printers are overly complicated. For example, too many resources are being diverted to carrying out repairs for individual errors that occur when converting CAD data to 3D data format. On top of this, insufficient compatibility between simulation and design formats is also increasing the number of steps and their complexity unnecessarily.

The conclusion we arrived at was that in order to address the issues above, it would be necessary to create a new 3D data format. The natural step was to look to the voxel. In the same way that pixels make up 2D images, voxels are the basic components and building elements of 3D objects. Additionally, the voxel exhibits great potential for future development and is an area where Fuji Xerox’s technological expertise in pixel-based image handling and processing has much relevance. Having decided on this direction, Fuji Xerox next approached Prof. Hiroya Tanaka of Keio University. A progressive figure in his field, Prof. Tanaka leads studies into how 3D technology can be made more accessible and contribute more to society. Hoping to draw from his wealth of 3D data processing knowledge, Fuji Xerox proposed to Prof. Tanaka the idea of joining hands to investigate the feasibility, applications, and potential of a new 3D data format.

Accelerated progress through common vision: “By making 3D printing technology available to the world, we want to provide solutions to the issues of today.”

Fujii: The objective of our joint research with Keio University was to study technology that would allow us to represent 3D models using voxels and to then be able to process that data. However, before deciding on any of these details, we first agreed on a common philosophy: “We want to enrich people’s lifestyles by making 3D printing available to the world.” This endeavor was not to be about the profit of any one entity, but about contributing to manufacturing worldwide, irrespective of field, and in doing so, providing real solutions to real issues in society today. Prof. Tanaka and I both felt very strongly about this.

The release of the FAV specifications in July 2016 was the first step towards realizing our goal, and steering us on this path was the core philosophy of our collaboration—our desire to bring solutions to society. Whenever any doubt came up over whether we were moving in the right direction, we always returned to these roots so that we could make the right decisions. Immediately after releasing the FAV specifications, we began receiving many inquiries and suggestions. We feel this signifies the potential of the FAV format and shows that our identification of the relevant issues was on target. It also makes us more firm in our belief that the societal issues which we hold important and try to solve through our collaboration are indeed relevant issues for which people require solutions.

Communication is key in identifying what is really necessary

Takahashi: In this joint research effort, the greatest challenge for us was ensuring that the team’s greater goal was properly represented in the technology and specifications. While the objective of the project was to propose a new 3D data format, we knew it had to be more than simply an improvement on existing 3D printer functions, and we also knew it needed to be a data format whose application would extend beyond just printing. So it was necessary for us to look ahead to the future when we carried out our investigation. This is where communication played a vital role.

In the field of 3D printing, there are several standard players: the printer and material manufacturers, designers, engineers, printer operators, etc. But also, outside of the manufacturing industry you have a wide range of users—those in architecture, medical care, nursing, education, etc.—and all of these players are expecting big things to come in the future of 3D printing. With this bigger picture in mind, we felt that the uses of the new 3D data format specifications and the benefits it could bring should not be limited to any one system, work process, or field. We then began to participate in consortiums and other events. At these events, we communicated with a number of leading 3D printing engineers and key figures in the manufacturing field, and what we found from our conversations with these people was that they all were experiencing struggles in creating accurate 3D data to produce the objects they envisioned.

In this project, we were able to gather together and organize the desires and expectations of a diverse range of people and make these a reality by defining them in the engineering requirements and specifications. I feel that this achievement was a true showcase of Fuji Xerox’s strength in drawing value from communication. Equally, it was the firm belief in this project held by Prof. Tanaka, myself, and all others involved that we were able to realize the data format we envisioned and that allowed us to release the specifications within no more than a year.

Specification release of the world-first FAV 3D printing data format

In 2016, we created and released the FAV 3D data format specifications to enable precise and complex expression through 3D printing. These specifications, the culmination of both Keio’s Prof. Hiroya Tanaka’s in-depth knowledge of 3D-data processing software development and Fuji Xerox’s image processing technology, were a world first in enabling the creation of three-dimensional objects capable of retaining complex internal structure information on colors, materials, bonding strength, etc. Furthermore, because FAV 3D objects are created by the arrangement of individual voxels (the basic elements of the FAV format), compatibility between simulations using finite elements is increased dramatically. This allows numerical analysis data from simulations to be used in design data as-is, removing the need for data conversion, and thus providing seamless movement between design and simulation environments. Furthermore, FAV specifications significantly increase simplicity in editing, making it possible to create and edit 3D data in a fashion similar to such tasks as cutting-and-pasting images and creating composite images. These and other such improvements in customizability are just some of the new developments that serve to further expand the possibilities of 3D printer manufacturing.

Specifications of the FAV 3D data format released here [PDF: 2.60MB]

See links below for technology details

Aiming to realize a new manufacturing environment utilizing 3D printers

In an aim to make FAV the de facto standard, we are currently implementing it in many activities, including the Japanese Ministry of Education, Culture, Sports, Science and Technology (MEXT) COI (Center of Innovation) project “Center of Kansei-oriented Digital Fabrication,” for which Keio University serves as a central base. Also, in 2016 we presented FAV at an international meeting held by ISO and ASTM International in order to make it the international standard. This in turn contributed to increasing Japan’s presence in the area of 3D data formats as well as the visibility of FAV.

In the future, Fuji Xerox plans to further enhance the specifications of the FAV format and provide handling technology for FAV data. Through promoting the effective utilization of 3D printers using FAV, we hope to realize a new manufacturing environment together with our customers.

Professor Hiroya Tanaka of Keio University

Professor Hiroya Tanaka Faculty of Environment and Information Studies, Keio University

At first, because most research projects related to 3D printers being conducted in Japan by the government or corporations tend to be devoted only to the physical aspects of 3D printing, such as equipment (machinery) and materials, I was strongly concerned that there may be a lack of recognition for the importance of its information-related aspects, such as 3D data and software. Our laboratory at Keio University had been working independently to develop platforms for 3D data and software, but we were searching for a partner company who could collaborate with us, working from a business-oriented perspective to make these platforms available and useful to society. It is then that we found Fuji Xerox. I feel truly fortunate that we were able not only to work together with Fuji Xerox on this technology transfer but also to share a greater vision with them. Going forward, I hope that we can continue to work together with Fuji Xerox to pursue the ideal of industry-academia collaboration. There are still many other challenges regarding 3D data that we need to work on besides the problems of expressing materials and colors, including flexible management of intellectual property rights and distributed management of data using blockchain technology.
In the near future, 3D data will also become linked with other areas such as gaming, virtual reality, and augmented reality. This is the dawn of an era in which value is created from data. The MEXT COI project for which our laboratory is responsible, and in which Fuji Xerox is also a participating member, is a long-term research project lasting until 2021. During the remaining time, it is both my dream and my goal to produce achievements of which Japan can be proud and transmit them to the world. Keio University also runs the database (, one of the world’s largest 3D databases that handles the FAV data format, and is also working on utilizing AI (artificial intelligence) and deep learning. Because this field of research is interconnected with many other areas, it can be troublesome, but I believe that this is also what makes it so fun.

  • Note 1 FAV (abbreviation for “fabricatable voxel”): a new voxel-based data format proposed by Fuji Xerox and Keio University in collaboration.
  • Note 2 The first 3D data format in the world to be able to retain not only information on the surface of a 3D model but also information on internal structures, colors, materials, and connection strength.