Collaborative 3D modeling system based on blockchain

TL;DR

This paper introduces a blockchain-based system for collaborative 3D modeling that enables decentralized version control and communication among users without relying on a central server.

Contribution

It presents a novel blockchain-based approach for decentralized, server-less collaborative 3D modeling and version control.

Findings

Decentralized mesh modification history stored on blockchain

Users can commit and checkout modifications without central servers

System enables collaborative modeling in a peer-to-peer manner

Abstract

We propose a collaborative 3D modeling system, which is based on the blockchain technology. Our approach uses the blockchain to communicate with modeling tools and to provide them a decentralized database of the mesh modification history. This approach also provides a server-less version control system: users can commit their modifications to the blockchain and checkout others' modifications from the blockchain. As a result, our system enables users to do collaborative modeling without any central server.

Figures

Full text

I Introduction

With the development of the game and animation industries to meet the desire and expectation of consumers, we have to deal with large and complex 3D models. To build large and complex models, collaboration among many designers becomes the common practice. Unfortunately, modeling tools like Blender does not support collaborative modeling yet due to its complexity. Instead, we can use a separate, version control system (VCS) such as Git, but there is a limitation that VCS depends on the central server(s) hosting the repository.

Departing from this centralized approach, the blockchain technology premises that we can share an immutable history of transactions without a central server. Based on the blockchain technology, we propose a new, collaborative 3D modeling system. Our approach has the following features:

• •

Decentralized, immutable, and Byzantine fault tolerant, because our approach is based on the blockchain.

• •

Modular, since the modeling part and the blockchain part of our approach are separated, so we can easily use various existing modeling tools.

We think that our approach is useful for server-less collaborative modeling projects, especially community based modeling projects. We expect our research to contribute to the modeling industry, and show that the blockchain can be used for various purposes even in graphics and modeling fields.

II Related work

Collaborative modeling. Doboš and Steed [2] presented an interactive visualization tool for mesh difference and conflict resolution. Also, a practical algorithm for diffing and merging polygonal meshes was presented [1]. Salvati et al. [5] proposed a real-time collaborative editing system of low-polygonal and subdivision meshes.

Blockchain-based revision control. There was a Kickstarter project [4], which tries to combine concepts of Git and blockchain. It attaches blockchain as a local proxy to Git for using Git without depending on Git hosting servers.

Blockchain-based database. A recent project attaches blockchain on a set of nodes where each node has its own MongoDB database to get the advantages of database and blockchain together [3].

While previous studies [1] [2] have focused on versioning meshes for collaboration modeling, our approach focuses on providing a distributed network that combines modeling tools with a decentralized database, allowing those prior studies to be used for server-less collaboration.

III Collaborative Modeling using Blockchain

In this section, we first talk about our blockchain network, followed by various operations that our approach support.

III-A Network

Our idea for collaborative 3d modeling is to regard the modification of the mesh as a transaction and use the blockchain as a database of modification history. In the network, each node consists of one or modeling tools, e.g., Blender and Maya, and a blockchain client (Fig. 1). The blockchain client is responsible for distributing the transactions to the network, managing the blockchain, and delivering the modifications from other nodes to the modeling tool.

One may consider to tightly integrate the modeling tool to the blockchain or vice versa. Instead of this integration approach, we choose the current approach of separately having modeling tool and blockchain client, since this decoupled approach enables users to use any existing modeling tool as a plugin or to use multiple modeling tools together.

III-B Operations

Our approach provides three operations: Commit, Mine, and Checkout.

Commit a mesh. When a user edits a mesh and requests commit, the modeling tool sends the mesh to the blockchain client. The blockchain client packs the mesh into a transaction. Instead of storing the whole mesh data, we store the difference between the current mesh and the prior mesh, which corresponds to the last transaction stored in the chain. We can retrieve the mesh later by iterating through the path from the root to the current transaction. This approach makes the transaction size a lot smaller than storing the whole mesh, so we can enable efficient collaborative modeling.

The blockchain client broadcasts the transaction to the peers, and each peer stores the transaction in a temporary storage, mempool, for the transactions waiting for being included in a block (Fig. 2).

Mine a block. When a user requests mine, the blockchain client packs the transactions in the mempool and mines a new block. It updates its chain and broadcasts the block to the peers. When the peers receive the block, they update their own chain (Fig. 3).

Checkout a mesh. When a user selects a block from the chain, and selects a transaction from the block and requests checkout, the blockchain client constructs the mesh corresponding to the transaction and send the mesh to the modeling tool. This operation provides user with a version control system. The user can download the latest commit, or rollback to one of the previous commits.

IV Implementation and Results

We have implemented our approach by using Blender and a simple PoW (Proof-of-Work) based blockchain written in Kotlin (Fig. 4). We used HTTP for implementing communication between Blender and the blockchain client, and communication between the blockchain clients. The blockchain client provides a simple GUI, which shows the mempool and the chain. The user can select a block from the chain and select a transaction from the block. When the user selects a transaction, the blockchain client provides 3D preview of mesh, which corresponds to the transaction. Our implementation can be run on multiple localhost nodes in a single computer, or multiple computers.

You can see the video of how our implementation works at https://youtu.be/s33wzSrMXwI. You can see the code of our implementation at https://github.com/Avantgarde95/C3DMB.