Boring log data tables, reconciled from field logs and lab sheets for the engineer of record.

Every boring leaves at least two records. The field log carries the boring number, depths, sample numbers, blow counts, water readings and the logger's descriptions. The lab sheet carries moisture contents, limits, gradations and strengths keyed to the same samples, sometimes weeks later and sometimes with a sample renamed. Before a report goes out, someone sets the two side by side, builds the tables and chases whatever doesn't match.

In the test, AI reads copies of both and drafts the report tables: boring, depth, sample number and blow counts exactly as recorded, lab values as reported, groundwater readings with their timing, and two lists, one of mismatches between log and lab sheet and one of anything missing. Each value points back to the page it came from. The draft classifies nothing and recommends nothing, and your reviewer compares it with the tables your firm issued.

Blow counts come in more than one dialect. Caltrans' 2022 logging manual has the field log record blows for each 6-inch increment, while its Log of Test Borings need not show them: it reports the second and third increments summed, writes a partial drive with its penetration to the nearest inch, such as 50/4 or 90/8, uses REF when the seating drive is never completed, and 0 when the sampler sinks under its own weight. North Carolina DOT's checklist for field staff, updated December 3, 2025, writes refusal as 60/0.0 or 60/0.1, with penetration in tenths of a foot, and reads 100 blows in under a foot one way in the Coastal Plain and another in the Piedmont. The table copies each count in the notation its log used and never converts one into the other.

Logs carry their own shorthand. The same North Carolina checklist records station and offset to the nearest foot and depths to the tenth of a foot, writes FIAD when a boring was filled in after drilling with no water reading, and wants sample numbers to run in sequence across the whole project with none repeated. Minnesota DOT's boring log terminology sheet abbreviates weight of hammer as WH, weight of rod as WR and no sample retrieved as NSR, and tags water readings AB, AC, AF or WSD for after bailing, after completion, after flushing, or while sampling or drilling. It also warns that its published logs are edited or abridged, with the field logs kept available at its offices. The draft keeps each code as written and names the key it came from.

The lab can overrule the field description. Caltrans names each sample by hole ID, a type code and a two-digit number counted from the top down, then a tube number, with U for a Shelby tube, S for a split spoon, B for bulk and C for core. Where a lab result differs from the field description and the geoprofessional judges it representative, the lab-based description goes on the log and into the report, and a test symbol beside the boring says so. The manual also wants hammer efficiency on the log, and one of three plain statements when groundwater was not encountered, encountered but not measured, or never measured. When a lab sheet's sample ID, depth or value doesn't match the log, the draft flags the pair and leaves the decision to your engineer.

Clients add rules of their own. Portage County's 2025 boring request wanted topsoil measured to the nearest half inch, any fill, debris or unnatural odor noted, coordinates and elevations delivered in .csv and shown on the final logs, and every boring that stopped short of its planned depth listed with the reason. Classification under ASTM D2487, the Unified Soil Classification System, leans on laboratory results, so the table carries the gradation and Atterberg limits beside each sample and leaves the classification to your engineer. The draft never assigns a group name or symbol.

What your reviewer will check. Does every sample in the table match its boring, depth and number on the log? Are blow counts, water readings and lab values copied exactly, in the source's own notation and units? Is every disagreement between log and lab flagged, every missing value listed, and nothing guessed? The tables in your sealed reports are the answer key. Classification, the soil profile and every recommendation stay with your engineer of record. None of this is legal or engineering advice.

Once your reviewer approves, new projects get the same draft tables for about 30 days.

Your engineer of record checks every table before it goes into a report, and each report is still signed and sealed the way it is now.

Live projects keep being compared with the tables your team builds. Any further job is a separate decision, priced on its own.