AI Requirements Platform for Construction Projects

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Building projects often face issues due to unclear objectives and contradictory requirements, leading to cost overruns, delays, and frequent change requests, as current planning software lacks the ability to analyze and automate holistic planning and design suggestions effectively.

Innovation Solution

A requirements engineering platform that integrates rational, functional, and emotional criteria into 2D and 3D planning, using intelligent questionnaires, natural language processing, and computer vision to derive automated planning recommendations, reducing the time and effort required from architects and planners to gather client requirements.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If architects and planners manually collect and process requirements from clients, then they can understand client needs, but the process becomes extremely time-consuming and expensive

Engineering Contradiction:
Improverequirements understandingVSAvoidtime for collecting and processing requirements
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The patent replaces the manual mechanical process of requirement collection with an AI-based automated system. The AI engine processes client inputs (text descriptions, images, floor plans) and automatically generates structured requirements, eliminating the need for time-consuming manual interviews and documentation while maintaining comprehensive requirements understanding.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables clients to independently provide requirements through self-service interfaces where they can upload images, describe their needs, and receive automated feedback. This reduces the burden on architects and planners while ensuring client requirements are captured accurately without manual intervention.

Inventive Principle:
Principle #25Self-service

2Adaptability or versatility

If architects and planners manually manage complex requirements, then they can handle contradictory needs, but the effort and cost increase significantly

Engineering Contradiction:
Improvehandling contradictory requirementsVSAvoidcomplexity of requirements management
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent replaces manual requirement management with an AI-based automated system that processes and resolves contradictory requirements. The AI engine analyzes multiple client inputs, identifies conflicts, and generates optimized design suggestions that balance competing demands, reducing the complexity burden on human professionals.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The AI engine acts as an intermediary between client requirements and architectural design. It mediates contradictory needs by analyzing all inputs systematically and generating balanced design recommendations, freeing architects from the complex task of manually reconciling conflicting requirements.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Ease of manufacture

If traditional 2D and 3D planning software are used, then architects can create design drawings, but the software lacks the ability to analyze requirements and provide automated holistic planning suggestions

Engineering Contradiction:
Improvedesign drawing creationVSAvoidautomated planning and design suggestions
Core Design Contradiction:
Ease of manufactureVSExtent of automation

Solution Approach 1:

The patent merges traditional CAD functionality with AI-based requirement analysis and automated design suggestion systems. The integrated platform combines the proven design drawing capabilities of 2D/3D software with automated requirement processing, enabling both manual design creation and automated holistic planning recommendations in a unified system.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent transforms specialized drawing software into a multi-functional platform that not only creates design drawings but also automatically analyzes requirements, processes client inputs, and generates holistic planning suggestions. This universal system handles both traditional design tasks and automated intelligence functions.

Inventive Principle:
Principle #6Universality (Multi-functionality)

4Reliability

If clear objectives and requirements are defined early in building projects, then project success improves, but current software lacks the capability to automate this process

Engineering Contradiction:
Improveproject success rateVSAvoidautomation of requirement analysis
Core Design Contradiction:
ReliabilityVSExtent of automation

Solution Approach 1:

The patent implements preliminary automated requirement analysis at the very beginning of the design process. The AI engine processes client inputs and generates structured requirements before detailed design work begins, ensuring clear objectives are established early while automating the entire requirement analysis process that previously required manual professional intervention.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP4478272A1Ai-based requirements engineering and recommendation platform for construction and renovation projects
Publication Date: 2024.12.18 PALASTS GMBH
  • EP4478272A1 patent drawingFigure 1
  • EP4478272A1 patent drawingFigure 2
  • EP4478272A1 patent drawingFigure 3

AI summary

Computer-implemented method for generating automated planning and design recommendations for buildings, the method comprising - implementing at least one computer vision model - implementing at least one natural language processing model - training the computer vision model(s) and/or language processing model(s) using special data sets - preparing at least one questionnaire containing questions based on the trained computer vision model(s) and/or language processing model(s) - presenting the questionnaire(s) to a user and recording his answers - generate recommendations and/or additional questions to the user