In the construction industry, quantity takeoff is one of the most important stages in cost estimation, tender preparation, and project cost control. Based on design drawings, engineers need to determine component counts, pipe lengths, finishing areas, and material quantities to prepare a Bill of Quantities (BOQ).
However, as the number of drawings increases, manual quantity takeoff becomes time-consuming, heavily dependent on individual expertise, and prone to errors when drawings are revised. This is one of the challenges that artificial intelligence (AI) can help address in construction.
By combining computer vision, AI-powered drawing recognition, optical character recognition (OCR), and domain-specific calculation rules, AI Quantity Takeoff can help identify objects, count components, measure dimensions, and extract structured data from construction drawings. When integrated with material databases, unit prices, and cost estimation workflows, the solution can support the automation of multiple stages of BOQ preparation.
How do AI Quantity Takeoff and AI BOQ work? Which tasks can be automated, and what should businesses consider when developing these solutions?
AI Quantity Takeoff refers to the application of artificial intelligence to assist with extracting quantities from construction drawings. Instead of requiring engineers to inspect every drawing and enter all measurements manually, an AI system can analyze drawing content, recognize trained or configured objects, and perform relevant quantity takeoff tasks.
Depending on the drawing type, input data, and implementation scope, the system can support the following functions:
Result consolidation: Export quantity data to schedules, CSV files, Excel spreadsheets, or formats designed for cost estimation workflows.
AI BOQ (Artificial Intelligence Bill of Quantities) refers to the application of AI to assist in preparing structured quantity schedules for construction cost estimation and tendering.
While Quantity Takeoff focuses primarily on determining quantities from drawings, AI BOQ can extend the workflow by organizing the extracted data according to work items, materials, units of measurement, and company-specific estimation rules.
An AI BOQ workflow may include:
It is important to distinguish quantity takeoff from completed cost estimation. To calculate estimated costs, the system also needs unit prices, applicable cost norms, work scope, and project-specific estimation rules.
A construction project may contain hundreds of architectural, structural, and mechanical, electrical, and plumbing (MEP) drawings. Engineers must inspect individual areas, cross-check symbols, measure dimensions, and consolidate data before completing the quantity schedule. As workloads increase, the time required for quantity takeoff can directly affect tender preparation, quotation deadlines, and construction planning.
Similar-looking symbols, drawings with multiple information layers, and inconsistent drafting conventions can cause engineers to overlook items, count the same object more than once, or enter incorrect figures. Even a small discrepancy during quantity takeoff can affect subsequent material calculations and cost estimates.
During design development, partition locations, doors, equipment, and construction scopes may change. These revisions can affect multiple related quantity items.
When data is managed manually, engineers must recheck drawings, recalculate quantities, and update schedules. The process becomes even more complicated when several drawing versions exist within the same project.
Different engineers may interpret drawings, apply measurement rules, and organize quantity schedules in different ways. When workflows depend too heavily on individual experience, businesses may struggle to standardize processes, train new employees, and maintain consistency across projects.
This creates an opportunity for AI to help transform quantity takeoff from a largely manual activity into a more structured workflow based on data, rules, and verifiable results.
An AI Quantity Takeoff system is more than a tool for reading drawings. To produce quantities that can be used in real construction workflows, the solution needs to combine visual recognition with technical data and calculation rules.
The system imports drawings in supported formats, such as PDF, DWG, DXF, or scanned images. Depending on the input data, it may use CAD parsers, image processing, OCR, or geometric recognition methods to extract the required information. For scanned drawings, image quality, resolution, scale, and symbol legibility can all affect recognition results.
AI analyzes shapes, symbols, and object characteristics to identify items within the configured scope. The important consideration is that the AI system must be designed for specific drawing types, symbol conventions, and business requirements. A single model should not be assumed to recognize every type of building component accurately without appropriate configuration or training.
After recognizing the objects, the system converts the results into data that can be measured and calculated.
Depending on the work item, calculation methods may include:
Not every quantity can be calculated through image recognition alone. Some calculations require dimensions, drawing scale, height information, material specifications, or domain-specific rules supplied by the business.
The extracted quantities are standardized and organized into a quantity schedule by work item, material type, unit of measurement, or building area. When unit prices and cost estimation rules are also integrated, the system can further support cost calculations, cost grouping, and exports based on company-specific BOQ templates.
AI should provide results that can be checked against the original drawings, allowing engineers to verify recognized objects, confirm quantities, and correct inaccurate entries. Items with low confidence scores, missing information, or exceptions to established rules should be flagged for review.
The goal is not to eliminate the role of engineers, but to reduce repetitive manual work so that they can focus on verification, technical judgment, and decision-making.
AI Quantity Takeoff can be developed for different construction workflows. Selecting the right initial use case helps businesses evaluate the benefits more clearly before expanding automation to additional work items.
AI can help identify rooms, measure floor areas, determine perimeters, and calculate quantities for walls, ceilings, skirting boards, and finishing materials according to predefined rules. This can support repetitive quantity calculations across multiple rooms, floors, and building areas.
AI can recognize and count symbols representing electrical equipment, fire protection devices, and other MEP components when the system has been appropriately configured or trained. Automating these repetitive counting tasks can help engineers consolidate equipment quantities more efficiently across large drawing sets.
The system can help identify pipe routes, cable runs, and other linear elements from drawings. By combining recognized geometry with scale information, dimensions, and classification rules, it can calculate and consolidate route lengths. The accuracy of these measurements depends on the quality of the source drawings and the availability of reliable scale and dimension data.
AI can assist with recognizing structural components, extracting dimensions, and consolidating data for selected beams, columns, and foundation elements. However, calculating concrete volumes, reinforcement quantities, and other complex structural items often requires information from multiple drawings, sections, reinforcement details, and discipline-specific measurement rules.
Extracted quantities can be linked to material catalogs, cost codes, unit prices, and company-specific estimation templates. This step transforms recognized drawing information into structured data that can support tender preparation, material procurement, and project cost control.
AI can assist with symbol counting, data extraction, and quantity schedule generation. Engineers can spend more time reviewing critical work items instead of performing every operation manually. The amount of time saved depends on drawing types, the number of items, the complexity of calculation rules, and input data quality.
When measurement rules, units, material codes, and table structures are configured within the system, businesses can reduce inconsistencies between employees and projects. This provides a foundation for workflows that can be repeated, measured, and improved over time.
When drawings are revised, the system can be designed to process the updated documents, compare quantities across versions, and highlight items that may have changed.
However, reliable change detection requires suitable version management and comparison mechanisms. Businesses should not assume that every revision will be detected automatically.
Standardized BOQ data can be reused in subsequent processes, including:
When connected to these systems, AI Quantity Takeoff can become part of an end-to-end digital construction workflow rather than an isolated drawing processing tool.
AI BOQ offers significant potential, but implementation results depend on drawing quality, quantity measurement rules, and the ability to verify outputs.
CAD drawings, scanned PDFs, and digitized paper drawings present different levels of complexity. Faint lines, missing dimensions, and inconsistent symbols can affect AI recognition accuracy. A solution should be evaluated against the actual drawing formats and conditions it is expected to process.
Each project may apply different standards, units of measurement, and quantity calculation rules. AI BOQ therefore needs to be configured around the company's actual business processes and technical requirements.
AI-generated quantities must be traceable to the source drawings so that engineers can inspect, identify errors, and make corrections before using the data. Verification is particularly important for complex items, ambiguous symbols, and calculations that require engineering judgment.
AI is generally more suitable for tasks with clearly defined rules, while complex cases or incomplete information may still require engineering expertise. Businesses should define the scope of automation carefully and maintain an appropriate human review process.
Instead of developing an AI BOQ system that attempts to cover every construction work item from the beginning, businesses can start with a focused research and development (R&D) initiative.
Step 1: Identify a priority use case: Select a frequently performed task or a specific drawing type, such as MEP equipment counting, interior finishing quantity takeoff, or pipe length measurement. A clearly defined use case makes it easier to establish measurable objectives and assess the feasibility of automation.
Step 2: Prepare data and calculation rules: Collect sample drawings, verified quantity takeoff results, material catalogs, and existing calculation rules. These resources provide the foundation for developing, testing, and validating the solution.
Step 3: Develop a prototype or proof of concept: Build an initial prototype to evaluate object recognition, quantity calculation accuracy, and data export capabilities. At this stage, the prototype should be treated as an experimental solution rather than a production-ready system.
Step 4: Evaluate performance using real project data: Compare AI-generated results with quantity schedules verified by engineers. Relevant evaluation metrics may include processing time, recognition accuracy, the number of manual corrections required, and the frequency of missed or duplicated objects.
Step 5: Integrate the solution into business workflows: Once the prototype meets the agreed evaluation criteria, expand the solution to include unit price data, BOQ templates, CAD/BIM integration, or project management systems as required.
Evaluation criteria should be defined before development begins. Examples include recognition accuracy, quantity deviations by work item, manual review time, and the ability to process previously unseen drawings. This approach enables businesses to evaluate AI based on measurable performance rather than demonstrations alone.
For AI Quantity Takeoff and AI BOQ projects, the challenge is not simply selecting an AI model. It is combining technology with technical data, quantity measurement rules, and the company's actual operational workflows.
With its Construction Domain × Software × AI approach, TGL Solutions focuses on developing software solutions tailored to business requirements, combining construction domain knowledge with software engineering capabilities. Depending on project scope, an AI Quantity Takeoff solution may be developed around the following functional areas:
These are potential development directions that can be customized to individual project requirements, rather than a claim that every function is already available in a single, fully commercialized product.
Businesses exploring AI for construction cost estimation can begin with a prototype focused on a priority workflow, evaluate its performance using real drawings, and gradually expand the scope before making a larger investment.
TGL Solutions can support the process from requirements analysis and workflow design to software development and technology integration, based on the agreed project scope. The objective is to turn a specific business challenge into a software solution that can be tested, measured, and further developed, rather than implementing AI solely for demonstration purposes.
If your business is exploring AI BOQ, automated construction quantity takeoff, or a custom AI module for CAD/BIM workflows, a focused R&D project or Proof of Concept (PoC) can be a practical starting point.
AI Quantity Takeoff and AI BOQ create opportunities to automate multiple stages of construction quantity takeoff and cost estimation. From symbol recognition and equipment counting to length and area measurement and BOQ generation, AI can help reduce repetitive manual work and improve data consistency. However, real-world performance depends on drawing quality, input data, business rules, and the ability to verify the results.
A practical approach is to begin with a clearly defined use case, develop a prototype or PoC, evaluate its performance using real project data, and gradually expand the scope of automation.
BIMCAD Vietnam develops BIM/CAD software, Web/Cloud platforms, and AI-powered solutions tailored to the construction industry. By combining domain knowledge, software engineering, and AI, we help businesses explore solutions aligned with their workflows and digital transformation goals.
Are you planning to develop AI Quantity Takeoff, AI BOQ, or a custom AI module for construction workflows? Contact BIMCAD Vietnam to discuss your business requirements, R&D scope, and the most suitable development approach.
