Starting the Loop
How L-SPARK can help in defence and dual-use
by Stewart Leckie
Director, Military, Defence & Dual-Use Technologies, L-SPARK
Commercial technology improves through a loop: build, put it in front of a customer, learn why it did or didn’t land, adjust, repeat. Between Canadian industry and the Canadian military, that loop has never turned. L-SPARK can start it. Accelerators already run founders through short, mentored cycles of testing and feedback. Pointed at defence, each cycle teaches a company the language and needs of the operator, and shows defence and primes what already exists in the ordinary economy. Companies learn their way to a fit that nobody could have specified in advance.
This piece builds on thinking Stewart first set out in “The Dual-Use Dimension,” published in 2067: A Journal of Canadian Consciousness and reprinted in The Walrus, and extends it to the role L-SPARK can play.
Canada’s $180 Billion Defence Buildup Is Already Out of Date
Canada is preparing to spend $180 billion on defence procurement over ten years, with a new agency to run the buying, ten named sovereign capabilities, and a target of 70 percent of purchases going to Canadian firms. The public debate is about the numbers. Stewart Leckie’s essay asks a different question: if the most valuable thing on offer is dual-use technology, is Canada structurally capable of taking delivery?
The argument rests on a reversal. For 50 years, dual use meant “spin-off”: the state funded the hard part for military reasons, and the civilian economy inherited the internet, GPS and radar. Canada has its own lineage here, from DATAR, the naval data-fusion system that worked on Lake Ontario in 1953, to the Canadarm. Today the dominant direction is “spin-in.” The frontier of machine learning, sensing, autonomy, batteries and space sits in commercial companies, and the decisive battlefield innovations in Ukraine and the Middle East have come from off-the-shelf parts and software shipped in weeks. Dual use now means one capability with two sets of customers and no conversion step between them.
That changes the state’s job. Under spin-off it was the origin point, which rewards patience, specification and control. Under spin-in it is the destination, which rewards speed, judgment and a tolerance for things built for other reasons. Canada’s institutions were built for the first job.
The deeper problem is that the market loop between Canadian industry and the Canadian military has never turned. Most technology firms have never had the Canadian Armed Forces in their addressable market; they were not rejected, they were simply never in it. Industry measures revenue and growth, while operators measure effectiveness, interoperability and sustainment. With no shared instrument, neither side can see the other, and there are almost certainly Canadian companies holding much of what the forces need without either party knowing.
The main door in, the request for proposal, can only buy what the institution already knows how to specify. Drawing on Ursula Franklin, Leckie argues that procurement is not broken but working as designed: predictable inputs, predictable outputs, never a breakthrough. Adding programs builds “a maze with excellent intentions.” The open question about the new Defence Investment Agency, staffed largely from the organizations it is meant to improve on, is whether its first intake process looks any different from the last.
He names six needs:
- A buying process that doesn’t outlast the company
- Clearances at the speed of the work
- A path that doesn’t force a choice between civilian and military markets
- Neutral ground for the parties to meet
- Capital shaped for ten-year hardware companies
- Enough published accountability that people come back.
The conclusion is cautiously hopeful. Spin-in has put the frontier within reach of a country with good universities, capable founders and functioning capital markets. What Canada lacks is “the habit of picking things up,” which comes down to small moves by procurement officers, investors, military officers, accelerators, primes and Indigenous rights holders. Otherwise, $180 billion will buy what we already knew how to ask for, from the people who already knew how to describe it.
