Plan-set QC for Texas highway projects
Catch the comments before the district does.
RoadQC reads every sheet of your submittal against TxDOT manuals, standard sheets and the 2024 spec book. It marks what a reviewer would flag, from design criteria to bid codes to typos, and cites the source for each one.
Example findings on a 60% plan and profile sheet
RDM-KCrest curve K = 61. A 60 mph design speed needs at least 151.
STD-REFNote cites TCP(2-1)-18, which is not bound in the set.
SPELL-TITLESheet title reads PLAN AND PROFLE.
PLACEHOLDERTitle block still reads SHEET XX OF 211.
Calibration, October 2026 · text-layer findings · details under Accuracy
How it works
From PDF to comment log in one pass
- Reading text layer173 sheets
- OCR on linework and scanned sheets38 sheets
- Matching the index to bound sheets211 sheets
- Checking 2024 bid codes and units642 items
- AI review of drawings211 sheets
Crest K = 61 at PVI 13+11; 60 mph needs 151.
TCP(2-1)-18 cited in a note but not bound.
Item 666 arrow marking bid in LF; this code is measured in EA.
| No. | Sheet | Comment | Reference | Response |
|---|---|---|---|---|
| 1 | 31 | Crest K = 61 at PVI 13+11 is below 151 for 60 mph. | TxDOT RDM, vertical curves | — |
| 2 | 31 | TCP(2-1)-18 cited but not bound in the set. | PS&E Prep. Manual | — |
| 3 | 7 | Item 666 arrow marking bid in LF; code is measured in EA. | 2024 bid codes | — |
| 4 | 2 | Index lists EC(1)-16; sheet 52 is EC(3)-16. | Index check | — |
What it checks
The comments reviewers write most, checked on every sheet
Each check cites the manual, section and edition behind it, so your designer knows what to fix and why.
Title sheet and index
Index against the sheets actually bound, spec book and FHWA-1273 dates, federal-aid number, reference markers, design-year ADT.
COMP-INDEXTITLE-SPECBOOKTITLE-FEDAIDStandards and references
Standard sheets cited in notes but not bound, citations without an edition, notes that point to sheets missing from the set.
STD-REFSTD-CITEXREF-TITLEBid items and quantities
2014 codes in a 2024 set, codes that do not exist, wrong units and descriptions, summary math, estimate against plan totals.
BID-CODEBID-UNITQTY-FORMULADesign criteria
Curve radius and K values against the design speed of each alignment, side slopes and PGL on typical sections, P&P scale pairs.
RDM-KTYP-SLOPEFMT-PP-SCALETraffic, bridge and drainage
MUTCD signing and work-zone values, bridge load rating and sign structures, SW3P forms and EC standards.
MUTCDBDM-LOAD-RATINGSW3PDrafting and spelling
Typos in titles and notes, street names spelled two ways, placeholders left in, missing seals, date stamps printing field codes.
SPELL-NOTECONS-STREETPLACEHOLDERBetween milestones
Know which review comments were really addressed
Compare the 30% and 60% sets. RoadQC matches sheets even when they are renumbered and follows each earlier comment into the new set.
| 30% comment | 60% sheet | Status |
|---|---|---|
| Typical sections show no side slopes or PGL | 4–5 | Addressed |
| Federal-aid project number left blank | 1 | Not addressed |
| General notes refer to illumination sheets not in the set | 6B | Verify |
| Layout sheet title misspelled | 3 | Addressed |
| No reference markers at project limits | 1 | Not addressed |
Accuracy
Measured on real plan sets, finding by finding
A QC tool that cries wolf costs you time. Every RoadQC finding in calibration was checked against the sheet by reviewers who had not seen the rules.
20 sets, four owners
TxDOT design submittals from 30% to final, ready-to-let sets, a toll authority and a city: 8,203 sheets in all.
Precision you can check
82% of text-layer findings were confirmed real. Most individual rules run at or near 100%; the rest are named and being tuned.
Recall against a district review
On a 30% submittal with the district's markup, RoadQC raised 64 of the 107 comments that could be checked from the plans.
Built for engineering firms
An extra reviewer, not a replacement for one
The engineer of record decides
RoadQC points to likely issues and cites its source. It does not review, approve or seal plans.
Your plans stay yours
Plan sets are never used to train AI models, and you choose how long they are kept.
Works with your owner's rules
TxDOT, toll-authority and city sets are checked against the requirements that apply to that owner.
Every check cites its source
Manual, section and edition on every finding, so a comment can be defended or dismissed quickly.
Pilot program
Run RoadQC on your next submittal
We are working with a small group of Texas transportation firms before general release.
- Findings on a live 30%, 60% or 90% set, with a walkthrough
- Direct input on the checks your reviewers want next
- Pilot pricing, locked in after release
Email us
Tell us your firm, your role and the type of projects you submit. We reply within one business day.