What This Series Demonstrated
Across the four preceding installments of The Ontario Roadmap, we traced a single thesis from the shop floor to the bid table.
In Part 1, a Hamilton machining shop accessed five-axis capacity it didn’t own — via an AI broker that matched its requirement against a Cambridge shop’s idle schedule, without forcing either firm to reveal its customer, its pricing, or its proprietary process parameters.
In Part 2, we extended that logic beyond the machine spindle: testing labs, specialized human expertise, and used equipment, all absorbed into the coordination mechanism, all resolving the kind of friction that leaves an independent SME unable to compete with a vertically integrated mega-factory.
In Part 3, we scaled it up. A small systems integrator assembled five independent Ontario firms into a dynamic consortium that outbid a Hegemon tier-1 factory on a hydrogen fuel cell contract — price, quality, and delivery time all competitive — using machines that already existed, talent already employed, and capacity that would otherwise have sat idle.
In Part 4, we asked the governance question: who owns the wire? The answer pointed toward an open protocol — Cosolvent — that allows competition among operators without allowing any single platform to capture the firms that depend on it.
This series established something real and actionable. AI-brokered flexible specialization, deployed across a region of independent manufacturing SMEs, can function as a virtual mega-factory. The Ontario Pocket is not a thought experiment. The architecture can be built.
But there is a limit to what The Ontario Roadmap addressed — and it is worth naming it precisely before we move on.
The Unit of Assembly
In every scenario in this series, the unit of cooperation was the whole firm.
Apex Milling entered an agreement with Tri-City Precision as corporate entities. Veridian Systems assembled five companies, each represented by its owner or operations director. The smart contracts, the NDAs, the margin splits — all of these ran between firms, authorized at the level of organizational decision-making, executed as commercial agreements.
This is appropriate. It is how business works at the scale of multi-million-dollar contracts. And the AI cooperation marketplace earns its value precisely by reducing the friction of those firm-level agreements enough to make them viable in the time frames that real manufacturing demands.
But consider what it cannot address.
A production engineer who needs forty hours of expert advice on a cutting parameter he has never seen before. A quality manager who needs a metrologist to program a CMM for three weeks, not three years. A machining shop that has a five-axis center running one shift when it could run two — not because of any contract opportunity, but because an organization forty kilometres away needs fifty machine-hours for a specific tooling job and has no way to find them. A maintenance supervisor at one firm who has deep experience with a specific PLC architecture, sitting idle on a Tuesday afternoon, while a technician at another firm is about to make a costly mistake on the same controller for lack of guidance.
These needs are real. They occur constantly. They cost the manufacturing sector in aggregate far more than the firm-level contract gaps that the Ontario Pocket mechanism addresses. But they are too small, too specific, and too episodic to be handled as firm-to-firm commercial agreements. Nobody writes a purchase order for forty hours of corridor-level expertise across a firm line.
The Problem Beneath the Surface
Underlying all of these smaller coordination failures is a structural feature of manufacturing that is so familiar it is rarely named: productive capability comes in lumps.
You cannot buy 0.4 of a five-axis machining centre. You cannot hire 30% of a metrologist. Floor space is leased by the unit, certifications maintained at a fixed overhead regardless of throughput, and specialized skills hired as whole people with full-time employment relationships. Every fixed asset in a manufacturing operation arrives as an indivisible unit that costs money whether you use it or not.
The inevitable result is that small firms seldom have exactly the right amount of any resource. They have too much when a contract ends and too little when the next one peaks. The idle five-axis machine, the underutilized specialist, the cramped floor in one corner and empty space in another — these are not symptoms of poor management. They are the predictable consequence of lumpy assets meeting variable demand in a business too small to absorb the variance internally.
At the individual firm level, this waste is unavoidable. At the level of the industry as a whole, it is a different story. One firm’s surplus is almost always another firm’s deficit. The idle shift on a Makino in Hamilton is the machine-time a Kitchener integrator needs next Tuesday. The metrologist who finished her qualification project last month is the expertise a Windsor medical device shop needs for the next six weeks. The surplus and the deficit exist simultaneously. The market they need does not.
Cooperative Specialization
The Ontario Roadmap described how AI-brokered cooperation resolves this problem at the firm level. The next stage of this architecture extends the same cooperation tools below the firm boundary — to departments, to individual machines, and to specific people working inside manufacturing organizations.
We call this stage Cooperative Specialization. The platform infrastructure that enables it — a Cooperative Specialization Support System (CSSS) — is the subject of the series that follows. It is, at its core, the same Cosolvent protocol that governs the Ontario Pocket — deployed at finer resolution, reaching inside the firm rather than only between firms. DeeperPoint’s MarketForge application is the intended prototype environment for demonstrating a CSSS deployment with real participating manufacturers.
A firm that connects to a CSSS can configure exactly which roles, which assets, and which domains of expertise are available for external engagement, under what conditions, and subject to whose approval. The cooperation is person-to-person or asset-to-asset in execution; the governance remains at the firm level.
What changes is the granularity of the market. Instead of a firm engaging another firm for a multi-month contract, a production engineer can be matched with a specialist practitioner for a week’s consultation. Instead of a firm sub-contracting a component program, a specific machine’s idle shift can be rented to produce a specific customer’s tooling. Instead of a firm hiring a fractional executive, a specific skill that lives inside one employee at one company can be offered to another company’s specific problem — authorized by both employers, tracked by the platform, governed by the same trust architecture that makes the Ontario Pocket contracts work.
The economic prize is substantial. Across Ontario’s manufacturing base, if even a fraction of the lumpy-asset mismatch waste that currently evaporates inside individual firms could be matched and redirected — idle machine shifts to firms that need them, specialist knowledge to problems that need it — the aggregate efficiency gain could rival the gains from the firm-level coordination we have already described.
What Comes Next
The Ontario Roadmap showed what becomes possible when individual firms learn to cooperate as a network. Cooperative Specialization is the vision of what becomes possible when that network reaches inside the firm — connecting people and assets across boundaries that commerce has historically kept opaque.
In the series that follows, we will explore this vision through concrete scenarios. Each one will begin with a familiar observation: a lumpy asset sitting below its potential utilization, a specific need going unmet because the right match is invisible. Each one will show what it looks like when the cooperation marketplace provides the match — not between companies, but between the people and assets that companies are made of.
(This concludes Series 3: The Ontario Roadmap. The next series — Cooperative Specialization — begins with an examination of the structural waste that Cooperative Specialization is designed to correct.)
What makes a thin market tick? → · The MarketForge platform → · The Cosolvent open protocol →