Canada is about to build at a scale it hasn't attempted in generations. The Major Projects Office is moving mines, transmission lines, ports, and northern corridors through coordinated approvals — the government talks about $200 billion in investment referred since last fall, with a stated path to $500 billion in private capital behind it.

Every one of those projects will generate something almost nobody is thinking about: a complete, outcome-linked record of how the physical world actually gets built. What the ground was like. What was attempted, with which machines and materials, under what conditions. What worked, what failed, what got reworked, and what it cost.

Right now, nearly all of that evaporates. It fragments across contractors, lives in proprietary systems, and dies on short retention schedules. We will spend hundreds of billions building the infrastructure — without a durable, shared record of the knowledge.

Norway solved this in 1995

There is a precedent, and it has been running for thirty years. Norway's Diskos National Data Repository requires petroleum licensees to report well data. The operator stays responsible for completeness and quality even when contractors create the data. Submissions carry structured metadata. Confidentiality windows protect commercial value before release.

Structured submission coexists with commercial rights — because reporting duties, quality standards, access rights, and confidentiality periods were designed together, not bolted on afterward.

Canada can do the same for the physical economy: mines, transmission corridors, ports, highways. Not by mandate — Norway's compulsory model doesn't transfer directly — but voluntarily, project by project, starting with one.

The proposal: one project proves it

A voluntary pilot. One willing proponent, one major project, eighteen months. Standardize the data the project already produces — work packages, inspections, equipment telemetry, material tests, environmental monitoring — into a common outcome-linked record. Govern it properly: tiered access, proponent keeps its IP, Indigenous co-design from day one under OCAP principles, Canadian-controlled infrastructure.

The valuable unit is not volume. It is an event linked to a verified outcome: what was attempted, under what conditions, with which tools and materials, what changed, and how the result was verified.

The honest framing, stated plainly: licensing is upside and cost offset, not a promise that data revenue repays a mine. The pilot tests construction-phase value — better regulatory evidence, faster root-cause analysis on defects and rework, and clearer safety corrective-action records — plus optional licensing revenue from assets whose collection was already paid for by construction. Maintenance and operating-performance benefits would need to be tested later.

Why a proponent says yes

Because the record can make construction delivery better: faster root-cause analysis on defects and rework, clearer inspection evidence, and better tracking of corrective actions. It can strengthen safety cases and regulatory relationships. And the pilot proponent helps develop a common record template that other projects could adopt if the pilot proves useful.

The first step is a bottom-up cost estimate, signed off by the proponent's own team, before any data flows. If the number doesn't work, they walk away having spent only the design phase.

We're looking for the first project

We've scoped this around the North Coast Transmission Line in BC as the reference design: $3.9 billion committed, active construction, a First Nations co-ownership opportunity. If you're building at similar scale in Canada — transmission, mining, ports, northern infrastructure — and you want your project to leave behind both assets, talk to us. One project proves the model.

During the design phase, the participants will develop and openly publish a common template for project records: what happened, under which conditions, how the result was verified, and who may access or reuse it. That template would be tested on the pilot before any wider adoption.

A shared standard for how Canada records its construction work should be developed openly with the people doing it. This proposal starts with one willing project—not a finished standard or a national policy.


Drew Clement is the founder of Archive Signal, which sources and brokers proprietary training-data corpora for AI builders.

The complete proposal

One project proves the model.

Read Two Assets: the voluntary pilot paper ↗

Sources & perspective

The investment context comes from Canada’s September 2026 announcement. Project and precedent sources are linked in the full paper.

This is an unfunded proposal, not an appointed pilot or a government program. Archive Signal has a commercial interest in the data layer. The proposed repository would be held by an independent steward, not Archive Signal.