InspectMind AI logo InspectMind

2026 Construction Drawing Issue Benchmarks

Explore 120 aggregate evidence pages across 15 recurring issue topics and eight design disciplines. No customer names, project names, addresses, drawings, or issue text are included.
By InspectMind··About 8 min read
Read the full error index

These pages make recurring construction-document patterns easier to compare without publishing a customer project as a marketing case study. Each cohort combines many completed reviews and reports only aggregate counts.

benchmark cohorts
120
completed reviews analyzed
1,114
visible findings analyzed
106,902
projects in the smallest cohort
27

Browse the benchmark library

Choose a topic, then compare it across architectural, structural, civil, MEP, fire-and-life-safety, and general coordination records.

Specifications, Schedules & References

References, schedules, notes, finishes, fixtures, and equipment information.

View topic findings and FAQs

Dimensions & Geometry

Dimensions, clearances, elevations, alignments, slopes, and geometric conflicts.

View topic findings and FAQs

MEP Coordination

Mechanical, electrical, plumbing, fire-protection, and cross-system coordination.

View topic findings and FAQs

Code Compliance

Building-code references, stated requirements, and compliance questions.

View topic findings and FAQs

Missing Information

Information described as missing, absent, omitted, unclear, or incomplete.

View topic findings and FAQs

Materials & Products

Materials, products, assemblies, finishes, insulation, coatings, and manufacturers.

View topic findings and FAQs

Structural Systems

Framing, foundations, slabs, reinforcement, loads, and structural connections.

View topic findings and FAQs

Building Envelope

Roofs, walls, flashing, waterproofing, moisture, insulation, and cladding.

View topic findings and FAQs

Doors & Windows

Doors, windows, glazing, frames, hardware, openings, and thresholds.

View topic findings and FAQs

Fire & Life Safety

Fire protection, smoke control, alarms, rated assemblies, exits, and egress.

View topic findings and FAQs

Civil & Site

Grading, drainage, utilities, stormwater, erosion, landscape, and paving.

View topic findings and FAQs

Accessibility

Accessible routes, ADA requirements, ramps, clearances, and maneuvering space.

View topic findings and FAQs

Occupancy & Egress

Occupant loads, exits, travel distances, corridors, stairs, and means of egress.

View topic findings and FAQs

Energy Performance

Thermal performance, energy codes, insulation, lighting controls, and efficiency.

View topic findings and FAQs

Document Status

Revisions, stamps, seals, signatures, dates, permit status, and addenda.

View topic findings and FAQs

Methodology and privacy

  1. Included projects were created from 2026-03-21 through 2026-09-16, had a completed review, and exclude deleted, internal, super-admin-owned, and test/demo/sample/untitled-prefix records.
  2. Dismissed, wrong-find, duplicate, insufficient-reference, and previously suppressed findings were excluded.
  3. Topics use disclosed, case-insensitive terms in finding titles. Discipline labels are normalized from review metadata.
  4. Every published cohort contains at least 10 projects and 25 findings.
  5. The published artifact contains aggregate counts only: never names, addresses, files, project identifiers, account identities, or finding text.

The September 16 refresh also excludes wrong finds and user dismissals. This changes the population from earlier editions; differences are not a growth or accuracy measure.

Download the aggregate data

Download all 120 published cohorts. The files contain only aggregate dates, thresholds, topic and discipline IDs, project counts, finding counts, and severity totals.

Limitations

Categories overlap: a finding can match more than one topic and can carry multiple discipline labels. Counts describe potential findings surfaced during InspectMind reviews, not confirmed defects, industry prevalence, or an error rate. Project scope, document size, design stage, and reviewer validation vary, so totals should not be used to rank disciplines by quality.