RESEARCH NOTE 001 / 15 OCTOBER 2026
A first Canadian industry baseline and a transparent manufacturing model. Test where a gain comes from—and what can erase it.
Analytical prototype The data describe existing industry. Scenario assumptions do not measure an achievable national gain.
NEW RESEARCH / 16 OCTOBER 2026
A recent Canadian project makes the energy question concrete. Inspect reported generation, construction costs and what continuous industrial loads require.
Source-backed scale arithmetic. National productive and household gains remain unestimated.
OBSERVED / CANADA, 2025
Billions of hours worked, across selected industries. These are baselines—not hours available for automation.
Source: Statistics Canada, Table 36-10-0713-01, retrieved 15 October 2026. Preliminary 2025 estimates. Mining includes oil and gas; construction includes more than housing. Selected industries do not describe the whole economy.
The table replaces archived Table 36-10-0480-01. A July 2026 correction restored aggregate jobs, hours, GDP and productivity, while detailed industry and compensation measures remained suppressed pending evaluation. Missing values are unknown, not zero.
We also found that the retrieved GDP and population series did not reconcile with the published per-capita indicator. That discrepancy remains unresolved. We removed per-person dollar projections and do not use that series to estimate household benefits.
ILLUSTRATIVE / CHANGE AN ASSUMPTION
Labour is one constraint. Materials, equipment, energy, quality and customers can bind first.
All scenario inputs are assumptions. This is a screening instrument, not a factory measurement, investment recommendation or national forecast.
METHOD / AN EXPLICIT PHYSICAL BOUNDARY
At a fixed workforce and product mix, relative hours per unit are q = (1 − f) + f / s + r. Here f is affected task share, s its speed multiplier, and r additional review labour per unit relative to its original labour requirement. The labour-limited gain is 1 / q − 1.
Realized output is capped by other capacity and demand. A zero headroom permits no positive extra output. Excessive review overhead can reduce capacity.
The central case assumes $600,000 upfront, $35,000 additional fixed cost/year and $8 contribution per extra unit before electricity. Additional power is 0.5 kWh/unit at $0.10/kWh. Benefits are halved in year one, over ten years at an 8% real discount rate, with zero terminal value.
Figures use illustrative CAD2026. Payroll is fixed; freed hours are not counted again as wage savings. Taxes, financing and unpriced worker/environmental effects remain outside this calculation.
| Case | Extra output | 10-year NPV, CAD2026 |
|---|---|---|
| Conservative assumptions | 3% | −685,859 |
| Central assumptions | 15% | −89,884 |
| Optimistic assumptions | 35% | +903,408 |
| Cheaper process improvement | 5% | +164,773 |
The process comparator assumes $50,000 upfront and $5,000/year. It is also uncalibrated. Equal-budget portfolios, real deployment constraints and spillovers require separate comparisons. Coordination must demonstrate value greater than its own cost.
REPORTED CASE / CANADIAN PRODUCTION
NGen reports that the Canadian Emergency Ventilators consortium produced 10,000 ventilators within 12 weeks. That implies an average of about 833 reported units per week across the effort. It is an aggregate production rate, not a measured unit cycle time or a before-and-after productivity gain.
Matched output and labour records, yield, full lifecycle costs, actual deliveries and an appropriate counterfactual. A later Health Canada recall concerning a sensor fault also makes quality part of the accounting. The reported ramp motivates investigation; it does not calibrate our illustrative model or establish the value of national coordination.
Sources: NGen’s StarFish case report and Health Canada’s April 2021 recall notice. Programme reporting and regulatory records answer different questions. No medical or engineering performance recommendation is made here.
FROM PRODUCTION TO PROSPERITY
A national output gain is not a matching increase in household income or wealth. Ownership, wages, taxes and transfers, foreign income claims, replacement of equipment, competition and the price of essentials affect who benefits. The research must measure these channels alongside output.
An inspectable industry baseline, reproducible conditional arithmetic and explicit bottlenecks. An independent AI reviewer reproduced the original model arithmetic and identified wording and source-scale issues; corrections and limitations are retained.
Achievable adoption, full sector capital needs, actual project returns, causal gains from national coordination and household distribution. No operating host, field demonstration or scientific validation is established.
SOURCES / REPRODUCE AND CHALLENGE
Prepared for GLEE with AI assistance from OpenAI Codex. Statistical observations, reported external cases, assumed scenarios and proposed research remain distinct. Source snapshot date: 15 October 2026.