Connected Cars X
Connected Cars X helps agents to set the smart coverage for aggressive drivers and help regular drivers to save on auto insurance. The solution consists of a Secure Telematics Hardware
Automated building plan approval system (ABPAS)
Our client, a German-based company, needed a building plan approval management system to automate the validation process and improve its speed.
Project background
In the modern German construction industry, in all building 3D projects, IFC schema specifications must be validated for compliance with current building regulations. The existing manual workflow takes a huge amount of time without the possibility to scale and speed it up.
The client needed a building plan approval management system to automate the process of validation, improve its speed, and shift the scaling process from increasing the number of employees to changing the technical capacity.
In order to automate updating of the legislation, we used NLP, Natural Language Processing. This technology makes it easy to extract rules from official documents, and fill and upgrade the regulation bases.
There are 2 roles in the building plan approval system: Client and Admin, where
Client (Construction company)
- Request creation (create a new building in the automated building plan approval system (ABPAS);
- Add IFC file (IFC file is a text version of a 3D model). It contains information about the building, rooms, and objects;
- Checking the building according to the specific rules. If the uploaded IFC files don’t match the rules, the app finds and highlights such files;
- Report with a detailed list of all inconsistencies.
Admin (Authorities)
- Rules management (add/edit/delete). A new rule requires a reference to the document that will initiate this rule. It can be either building regulations or buying laws;
- Authority check;
- Report (list of inconsistencies);
- Statistics of all requests.
- Feb 2020 – April 2020
- Germany
- Construction
- .NET Azure Application Insights
- Azure Cosmos DB
- Azure Storage
- Azure Active Directory
- Pandas
- Numpy
- Stanza
- Sklearn
- Python
- React.js
- Xbim Toolkit
- Docker
- Azure DevOps
- 3D models processing Fast uploading and accurate processing of 3D building models
- Compliance check There is an option to check the compliance on the authorities' side
- Verification Verification against building laws and regulations
- Feedback Getting transparent feedback on the building model
- Permit status check There is an option to check the status of the building permit
Business needs
Our client needed a building plan approval management system to automate the process of validation, improve its speed, and shift the scaling process from increasing the number of employees to changing the technical capacity.
- Reducing the time needed for considering and decision making;
- Speeding up the validation process;
- Decreasing the number of inaccurate 3D models that do not meet building regulations;
- MVP product development within 3 months.
Solution
CHI Software has won the tender announced by the authorities. The automated building plan approval system (ABPAS) allows uploading 3D models of the building and automatically validating its compliance with local regulations. The validating process narrows down to 4 steps:
-
A construction company uploads an IFC file (3D model) — the model of a building.
-
The building plan approval management system verifies the project against existing regulations (the building construction rules are set to the system by municipal officials).
-
Officials study the 3D file on their side, check the compliance, ask clarifying questions using the provided survey form.
-
The company gets transparent feedback to correct mistakes quickly.
Client values
- We removed most of the manual steps needed for verification.
- Significantly simplified the interaction between construction agencies and authorities.
- Released a stable MVP version within the agreed time.
Employee testimonial
The building plan approval management system, developed by CHI Software, has removed most of the manual steps in the validation process and simplified the interaction between construction agencies and municipality officials. Thanks to Natural Language Processing (NLP) technology, it became possible to extract the rules with high accuracy. Moreover, the application works with IFC building models. And the NLP module was designed to translate building rules into requests to the IFC class. Finally, the project validation for compliance with regulations has been reduced to just a few clicks.
Alexandr Kolosov ML Technical Lead