The Fork: Why the Same Dataset Has Two Futures
Every spatial capture workflow — every laser sweep, every photogrammetric pass, every orbit — arrives at the same quiet junction. Almost nobody notices it. It is the most expensive unnoticed moment in the built world.
Watch a capture crew work and you will see enormous care taken with the front of the workflow. The instrument is calibrated. The targets are set. The registration is checked to fractions of a millimeter. Then the dataset — millions of points, each one a small act of measurement — is exported to a format like E57 or LAS, and the care abruptly ends. The file enters what we call the terminal workflow: it is modeled, analyzed, flattened into a report, and shelved.
None of this is malpractice. Scan-to-BIM is real work. Deviation analysis answers real questions. The report that lands on a project manager's desk is genuinely valuable — once. The problem is what the terminal workflow does to the dataset the moment that value is delivered: it stops treating it as data. The file becomes an attachment. The attachment becomes an archive. The archive becomes a liability with a storage bill.
PATH A · T+90 DAYS → VALUE DECAYING — TRUTH GOES STALE
PATH A · T+5 YEARS → VALUE STORAGE COST · NEXT QUERY IMPOSSIBLE
What the export formats cannot say
Here is the deeper issue, and it is structural, not procedural. Every standard export format in the capture industry — E57, LAS, LAZ, PLY, OBJ, GeoTIFF, STEP, all of them — carries geometry beautifully and carries provenance not at all. Open any of those files and you can learn where every point sits in space. You cannot learn, in any governed and verifiable way, which instrument produced it, under what calibration, with what accumulated uncertainty, anchored to which datum, captured by whom, and how it has changed since. The formats describe shape. They cannot testify.
That missing testimony is precisely what institutions need. A lender pricing risk against a physical asset, an insurer underwriting its condition, an agency certifying its compliance — none of them can act on shape alone. They act on trust, and trust requires a record that can answer for itself. The capture industry produces truth every day and exports it into formats that cannot hold onto it.
Path A sells the answer.
Path B owns the instrument.
The other branch
The SDTX Exchange workflow begins at the same junction with the same file — and does something categorically different with it. The dataset is canonicalized: its geometry is fused with the metrological record the export format dropped — instrument, calibration, uncertainty budget, datum chain, operator credential — and sealed under SD&T notation, nine compartments of governed truth. The result is a .sdtx record: not a description of an asset, but an instrument denominated in it.
From there the record enters custody — hash-linked into an append-only vault — and receives a score: D_SDTV, one dimensionless number that means the same thing at nanometer and kilometer scale. A number an underwriter can price. A number an API can serve. And because the physical world never stops moving, the record never stops working: the asset drifts, the score decays, and physics itself mandates the recapture that renews the record, deepens its chain, and triggers the next cycle of value.
PATH B · T+300 DAYS → DRIFT DETECTED · RETRIGGER FIRED
PATH B · CYCLE 2 → .SDTX №00046 · CHAIN DEPTH +1 · VALUE COMPOUNDING
Why the fork matters more than the capture
The capture industry has spent two decades perfecting the front of the workflow — better lasers, denser clouds, faster registration. All of it feeds the same junction, and nearly all of it flows down the terminal branch. The opportunity is not to capture better. The opportunity is to change what capture becomes. The same crew, the same instrument, the same field day — routed through the exchange branch — produces an asset that appreciates instead of an attachment that decays.
That is the whole thesis, and it fits in a sentence: the value was never the capture. It is the life of the file.