Automated Bill of Quantities: How Drawings Become Quantities

An automated bill of quantities is a bill of quantities generated by software that reads construction drawings and measures the elements in them, instead of an estimator scaling, counting and typing by hand. For structural work, that means taking DWG, PDF or IFC files and returning structural steel and rebar in tonnes and concrete in cubic metres, with each line traceable to the drawing it came from. Speed is the obvious benefit, but the real test of an automated bill is whether it shows its assumptions.

What is an automated bill of quantities?

An automated bill of quantities is a list of measured work items, organised by trade chapter and unit, that software extracts directly from the drawings. The output looks like a conventional bill of quantities (BOQ): item descriptions, units and quantities that an estimator can price, compare with a tender document or hand to a client.

The difference lies in how the numbers are produced. In a manual takeoff, an estimator reads every plan, section and detail, measures lengths and areas, counts repeated elements and enters the results into a spreadsheet. In an automated workflow, the software does the reading and measuring, and the estimator's role shifts from producing numbers to reviewing them.

Three chapters of structural work suit this approach particularly well, because their quantities follow directly from geometry and standard reference values:

All three chapters share one feature: the arithmetic is simple, but the inputs are scattered. A slab thickness may sit in a section, a bar diameter in a schedule, a steel profile in a legend. Collecting those inputs correctly is most of the work, whether a person or software does it.

How does software turn a drawing into a bill of quantities?

Software turns a drawing into a bill of quantities by identifying structural elements, reading their dimensions and specifications, and converting the measured geometry into billing units chapter by chapter. The steps mirror what a careful estimator does; the difference is that they run the same way on every sheet.

1. Ingest the set. The drawings are uploaded as DWG, PDF or IFC files, or as a ZIP of the whole set. A complete set (plans, sections, elevations and typical details) matters, because a dimension missing from a plan is often found on a section or in a detail.
2. Choose the chapters. The user selects which chapters to measure, for example rebar and concrete only, before the run starts.
3. Identify elements. Beams, slabs, columns, footings, piles, steel members and bar callouts are recognised from geometry, text, legends and notes.
4. Resolve dimensions. Each element needs a complete set of values: a slab needs a thickness, a bar needs a diameter, a count and a length, a steel member needs a profile and a length.
5. Flag what is unclear. Where a value is missing or two sheets contradict each other, the question goes to the user instead of being filled in silently.
6. Calculate and summarise. Lengths, areas and volumes become tonnes and cubic metres, grouped by chapter, element type and diameter or profile.
7. Link back to the source. Each line keeps a reference to where it was measured, so a reviewer can check it against the drawing.

Step 5 is where tools differ most, and it is where most of the financial risk sits. A tool that fills every gap with a default value will always produce a complete-looking bill, which is exactly what makes its errors hard to see.

What does an automated takeoff measure in steel, rebar and concrete?

An automated takeoff of structural work measures three quantities: the weight of structural steel, the weight of reinforcing steel and the volume of cast concrete. Each chapter has its own calculation logic and its own typical traps.

Structural steel

Structural steel is measured as the length of each profile multiplied by its mass per metre from standard tables, plus plates and connection material. An IPE 300, for example, weighs 42.2 kg/m and an HEB 300 weighs 117 kg/m, so confusing two profiles changes the weight of a member by a large factor. The common trap is a member whose profile differs between the plan and the section. The full method, with a table of common profiles, is in the guide to structural steel quantities.

Rebar

Rebar is measured as total bar length per diameter, converted to weight, because the mass of a bar grows with the square of its diameter. A metre of Ø32 weighs 6.313 kg, about sixteen times a metre of Ø8 at 0.395 kg. Laps, hooks and bends add length that is easy to leave out. The weight table and a worked beam example are in rebar quantity calculation.

Concrete

Concrete is measured as volume: length × width × depth for slabs, beams, columns and footings, and π × r² × h for circular piles. A pile of Ø60 cm, for example, holds about 0.283 m³ of concrete per metre of length. The thickness of a slab is frequently not written on the plan at all but in a section, a general note or a legend. The formulas for each element are in concrete quantity calculation.

Manual takeoff, generic software or AI that reads drawings: what is the difference?

The difference between manual takeoff, generic measuring software and AI that reads drawings is who does the reading: a person with a scale rule, a person with a digital measuring tool, or software that interprets the drawing itself. All three can produce a correct bill of quantities; they differ in effort, consistency and how visible the assumptions are.

Generic measuring software speeds up the measuring step, but someone still has to decide what each line represents and convert lengths into weight. AI that reads drawings attempts the reading step as well, which makes the quality of its questions and the traceability of its results the main things to evaluate. A fuller comparison, including what to test in a demo, is in the guide to bill of quantities software.

Why are hidden assumptions the main risk in an automated BOQ?

Hidden assumptions are the main risk in an automated bill of quantities because a number built on a guessed dimension looks exactly like a correct one. Speed makes the problem worse: the faster a bill appears, the less likely anyone is to search it for the guess.

Three silent errors account for most of the damage:

• Assumed dimensions. A slab thickness that is not stated on the plan gets a default value. Illustrative example: on a 100 m² slab, assuming 200 mm when the section shows 250 mm leaves out 5 m³ of concrete, a fifth of the true volume.
• Double counting. The same beam appears on the floor plan and on a detail sheet and is measured twice, along with its reinforcement.
• Elements that fall out. Piles or reinforcement cages drawn only on a foundation sheet or in a schedule never make it into the bill.

None of these errors produces a warning on its own. The safeguard is a process that refuses to fill gaps silently. Quantap is built on that principle: Quantap doesn't guess, it asks. When something in the drawing is unclear, missing or contradictory, it asks the user through Quick Check, and the quantities update with the answers. The article on hidden assumptions in a bill of quantities explains how to detect each of the three errors before submission.

How should you review an automated bill of quantities?

An automated bill of quantities should be reviewed the way a senior estimator reviews a junior colleague's takeoff: by sampling lines back to the drawing, checking totals against expectations and confirming every assumption. The review takes far less effort than producing the bill by hand, but it should never be skipped.

A practical review covers seven points:

1. Units against the contract. Confirm that rebar and steel are in tonnes, concrete in m³, and that the item breakdown matches the tender format.
2. Traceability. Pick lines at random and trace each one to the sheet and element it came from. In Quantap, every line in the report can be checked back against the drawing.
3. Answers to open questions. Re-read every answer given during Quick Check. One wrong answer about a typical slab thickness affects every floor that uses it.
4. Completeness. Compare the element list against the foundation plan, schedules and typical details to catch items that fell out.
5. Duplicates. Look for elements that appear on more than one sheet, such as beams shown on both a plan and a section.
6. Revisions. Confirm that the latest revision of each sheet was used, and that superseded sheets were not uploaded alongside it.
7. Sense checks. Compare ratios, such as kilograms of rebar per cubic metre of concrete, with what is typical for the structure type in your market.

Automated quantities are for professional review and do not replace professional engineering judgement. Responsibility for a submitted bill stays with the estimator and engineer who stand behind it. The complete pre-submission list is in the bill of quantities checklist.

Which drawing formats work best for an automated bill of quantities?

DWG, PDF and IFC all work for an automated bill of quantities, but each format carries different information. DWG files keep geometry, layers and text as editable data; PDF files preserve the issued drawing as printed, sometimes with vector geometry and sometimes only as an image; IFC models describe building elements with properties rather than as lines.

Format matters less than completeness. A single plan in the best format still lacks the section that states the slab thickness or the typical detail that defines a stirrup. Quantap reads DWG, PDF and IFC files and accepts a ZIP of the whole set, and the recommendation is always to upload plans, sections, elevations and typical details together. The trade-offs of each format are covered in DWG vs PDF vs IFC for quantities, and common questions about uploads and results are answered in the FAQ.

Summary

An automated bill of quantities replaces manual measuring with software that reads drawings and returns structural steel and rebar in tonnes and concrete in cubic metres. The quality of the result depends on two things: a complete drawing set and a process that surfaces unclear items instead of guessing. Assumed dimensions, double counting and missing elements are the errors that matter most, and none of them announces itself. Treated as a fast, traceable first pass that a professional reviews, automation changes where an estimator's time goes rather than whether the estimator is needed.