HomeLab.ai

This project is a special combination of many different technologies including AI and Blockchain to create products that can be applied in practice with many functions related to automating drawing design in construction.

Project's goals

• By automating the process, the system shortens the time help user quickly create 2D design and hire designer.

• Applying new technology to make a new way to building house with cheaper method.

• Using blockchain at the system to transparent the transaction process.

HomeLab.ai home page

Main functions

• 2D design analysis with TF2 Floor Plan (AI service)

• Design the building function diagram using G2P (AI Server)

• Export 2D designs to popular image formats

• Feature extraction of the following passage describing the house

• Make Order

• Utilize smart contracts to contract a designer.

• Put drawings on the marketplace

• View 3D drawings directly on the app

• Use AI to assign and recommend home furnishings

• Estimated cost for furniture

• Offer interior designs for the specified home frame.

Customers input their house's idea

AI describes characteristic of the customer's house

AI model will detect and recommend separate room areas based on overlapping color histograms

AI suggest floorplan diagram

2D drawing is suggested and implemented by the AI model

3D model after built

3D model allows viewing from different angles

Building materials inside the product are traced using the blockchain

Order description

Extracted floorplan drawing by AI

Designers can post the products on HomeLab.ai, we help identify the products by blockchain

Materials identifier area

Core technologies

We introduce a learning framework for automated floorplan generation which combines generative modeling using deep neural networks and user in-the-loop designs to enable human users to provide sparse design constraints. Such constraints are represented by a layout graph. The core component of our learning framework is a deep neural network, Graph2Plan, which converts a layout graph, along with a building boundary, into afloorplan that fulfills both the layout and boundary constraints. Given an inputbuildingboundary, we allow a user to specify room counts and other layout constraints, whichare used to retrieve a set of floorplans, with their associated layout graphs, from a database. For each retrieved layout graph, along with the input boundary, Graph2Plan first generates a corresponding raster floorplan image, and then a refined set of boxes representing the rooms. Graph2Plan is trained on RPLAN, a large-scale dataset consisting of 80K annotated floorplans. The network is mainly based on convolutional processing over both the layout graph, via a graph neural network (GNN),and the input building boundary, as well as the raster floorplan images, via conventional image convolution. We demonstrate the quality and versatility of our floorplan generation framework in terms of its ability to cater to different user inputs. We conduct both qualitative and quantitative evaluations, ablation studies, and comparisons with state-of-the-art approaches.

Architecture of our Graph2Plan network

The authentication system implemented on the Binance Smart Chain (BSC) blockchain network is currently in wide use to ensure the stability of the blockchain network. In order to ensure high reliability for Smart contract, the system only allows material validators that the system has previously authorized to participate in material validation.

Blockchain and Smart Contracts in Construction

• Step 1: The user uploads a contract to the smart contract with the content of the house idea, including the 2D design and extra details. The contract compels the designer to develop and complete the architectural and interior design within the time frame indicated.

• Step 2. After reading the contract, the designer determines whether to accept or reject it by submitting the transaction to smart contract.

• Step 3. After reaching an agreement, the designer creates at least three designs and uploads them to smart contract.

• Step 4. Users assess and select the best design, or they can request that the designer rework it. Each time the user makes a new request, he or she must pay a compensation fee equivalent to 10% of the incentive placed in the smart contract in step 1.

• Step 5. The work is repeated (up to ten times) until the user accepts the best design and the contract is closed.

• Step 6. The designer receives a bonus for the contract's termination

System architecture

We need a decent design in order to create a high-quality product. Users will interface with the business through BE Server since it is set up in a centralized manner, rather than directly with its functionalities (also known as intermediate server).

Model describing a server cluster