Reflections from ROAR East 2026 – York University, Toronto
There is a version of the Canadian defence robotics story that sounds like a press release. Increased NATO spending commitments. Dual-use technology opportunities. A booming domestic UAV and UGV sector ready to supply both the Canadian Armed Forces and the first responder community. A generation of entrepreneurs finally permitted to take defence seriously.
That version is true. It is also incomplete.
We recently attended ROAR East – the Remote Operated and Autonomous Robotics Conference– held at York University’s Schulich School of Business in Toronto. ROAR is Canada’s premier conference series dedicated to unmanned systems, drones, and autonomous technologies, bringing together buyers, operators, manufacturers, and policy makers across public safety, defence, energy, mining, and government. This particular edition was one of the more concentrated gatherings of people actually building, buying, deploying, and funding defence and dual-use robotics in Canada. The room included serving and former military leaders, drone company founders, first responder units, counter-UAS specialists, venture investors, a Ukrainian battlefield adviser, and a Croatian drone manufacturer who has shipped hardware to active conflict zones.
What emerged over two days was a more complicated, more useful picture than the press release version. One with structural implications for any hardware company – or manufacturing partner – trying to serve this market seriously.
The Battlefield Has Already Moved
The most important framing came from General Wayne Eyre, former Commander of the Canadian Armed Forces.
His message was precise: the military has a long history of being fascinated by new technology and then struggling to turn it into battlefield advantage. Robotics will be no different unless development is disciplined around operational need rather than technological novelty.
He identified the mission areas where unmanned systems can make an immediate difference – logistics, casualty evacuation, ISR, explosive ordnance disposal, perimeter security, urban reconnaissance, maritime surveillance. For Canada specifically, he noted that our vast geography and relatively small population density make robotics not just useful but strategically necessary. We need systems that can operate in harsh, remote, and contested environments.
Then he made a point that should land hard for anyone designing or manufacturing these systems: a platform that performs well on a demonstration range but fails when GPS is denied, communications are jammed, batteries are cold, or mud clogs its tracks is not operational capability. It is a trade show exhibit.
This is not a peripheral concern. It is the core design requirement.
Ukraine reinforced this. Olena Kryzhanivska, who advises on Ukrainian battlefield innovation, described the pace of change on the ground – hardware iteration cycles of four to six weeks, software cycles as short as seven days, and frontline R&D labs making final modifications to equipment as close to the line of contact as safely possible. She described how units have moved from experimentation to large-scale deployment of unmanned ground vehicles for logistics, engineering, and fire support, and how the forces that adopted most successfully were those that stayed close to operators, iterated continuously, and treated the operational environment as the real test bed.
She was also clear about what international observers consistently misunderstand: autonomy is not replacing the human. Drone operators, crews, engineers, and support personnel are the system. The hardware is only as effective as the people trained to use and maintain it – and the integration of that hardware into broader combined arms doctrine.
General Eyre made the same point from a military strategy perspective. The lesson from Ukraine is not that numbers alone win. It is integration – systems that plug into the broader fight, reduce human exposure to the most dangerous tasks, and allow commanders to generate effects more precisely.
Dual Use Is Not a Feature. It Is the Business Model.
One of the clearest strategic themes across both the defence and public safety streams was the value – and the commercial logic – of dual-use positioning.
Kevin Todrel‘s opening keynote laid out the structural shift: drones originated as hobbyist technology, moved into commercial and industrial applications, and have now pivoted hard into defence. That pivot brought significantly more funding, more rigorous technical requirements, and a faster iteration cycle. The result is a class of technology that is genuinely useful across forestry, infrastructure, natural resource extraction, search and rescue, ISR, logistics, and frontline operations – often with the same or very similar hardware.
From an investor’s perspective, this matters enormously. One panellist at the Canadian ecosystem session made the point directly: a dual-use product in search and rescue or infrastructure inspection has commercial customers who pay today, which means a supplier serving a defence contract that runs three months late does not go broke waiting. There are other places to sell, other places to test, and other revenue streams to sustain the company.
From a product development and manufacturing perspective, dual-use positioning creates different but equally important pressure. A system designed only for military procurement can be over-engineered for its commercial equivalent, or under-designed for actual operational conditions, because neither requirement fully disciplines the architecture. Systems that must perform across both domains – RCMP drone docking stations, DFR platforms, ISR drones used in both public safety and defence ISR roles – tend to force better engineering decisions earlier.
The drone-as-first-responder deployments discussed in the public safety stream illustrated this concretely. Durham Region Police, RCMP, and others described response times dropping from twenty to thirty minutes to under four minutes through pre-positioned docking stations and remote piloting. Launch times as low as thirty-five seconds. Drones on-scene before responding officers, transmitting thermal imaging in real time, coordinating arrests, locating missing persons in dangerous conditions. These are not military missions, but the hardware requirements – reliability in cold weather, GPS accuracy, low-light imaging, communications resilience, autonomous return-to base – are effectively the same.
The Supply Chain Problem Is Structural, Not Logistical
Across multiple panels, the conversation kept returning to supply chain – and what was said was more honest than the usual talking points.
Ivan Jelusic, CEO of ORQA FPV, gave the most direct account. His company started building FPV racing goggles. When Russia invaded Ukraine in early 2022, he shipped thirty drones in a diplomatic pouch because someone he knew needed them. That moment became the pivot point for ORQA’s transformation into a Western defence supply chain company.
His diagnosis of why the West cannot simply compete with DJI on technology features was blunt: DJI dominates because it owns its entire technology stack, from propellers to motors to flight controller firmware to radio systems. System integrators assembling products from off-the-shelf components cannot reach that level of performance or cost efficiency. Western companies that want to compete cannot chase DJI’s technical superiority – they will lose. What they can do is own the software and AI layer, build sovereign manufacturing capabilities, and compete on compliance, reliability, and interoperability rather than on raw specification.
He described ORQA’s current strategy in Canada as building exactly that: a manufacturing hub that enables Western companies to build drone solutions using a supply chain independent from Chinese mission-critical components, with sovereign software and AI capabilities developed locally.
That last point echoes what Canadian drone manufacturers on the CEO panel said about their own supply chain decisions. One company described walking across the street to a GTA manufacturer rather than sourcing from overseas, designing for component interchangeability from the start, and investing in in-house sewing capacity, CNC equipment, and injection moulding specifically to reduce dependency and enable rapid design changes without waiting on external suppliers. When COVID hit, that vertical integration let them navigate chip shortages and supplier disruptions better than their competitors.
The underlying principle is the same in both cases: supply chain resilience is not a procurement problem. It is a product architecture decision. It is made – or not made – during design.
The Gap Between Prototype and Fielded System Is Where Things Break
Across all the sessions, a consistent failure pattern emerged – one that defence procurement veterans, battlefield advisers, and venture investors described from different angles but arrived at the same diagnosis.
Companies build capable prototypes. They demonstrate them at conferences, submit them to procurement processes, and sometimes win contracts. Then the hard part begins: producing them reliably, at volume, in configurations that survive operational environments, with supply chains that do not collapse under pressure, and with the ability to iterate quickly based on user feedback.
Most companies are not structurally prepared for that transition.
Olena Kryzhanivska described Ukrainian forces rejecting a supplier not because the product was technically inadequate, but because the company was not flexible enough to iterate when field conditions changed. The product team believed their solution was finished. The military unit moved on.
General Eyre described the same dynamic from a procurement perspective. The speed at which Ukrainian frontline units are incorporating field feedback into production – ideas from the front line appearing in the production line within days – is the standard that matters. Canada’s procurement processes are not designed for that. But hardware companies that want to serve serious defence buyers have to be able to move at something closer to that tempo, at least in their development and manufacturing processes.
Investors on the funding panel raised the parallel problem from a capital perspective. Canadian companies that win early defence-adjacent contracts – grants, pilot programs, government trials – often hit a ceiling when it comes to scaling. Either they lack the manufacturing capacity to fulfil larger orders, or their product has not been built for volume production, or their supply chain depends on single sources that cannot ramp. The “missing middle” between early-stage demonstration and scalable production is where many companies stall.
This is not primarily a funding problem. It is a product development and manufacturing readiness problem that appears as a funding problem.
What Investors Actually Said – and What They Are Not Going to Say Publicly
The funding panel at ROAR – which included Devon Galloway of Garage Capital, Matt Lombardi of On9 Ventures, and Michael Langlais of Canadian Corps of Commissionaires – was more candid than these conversations usually are. A few things said in that room are worth repeating outside it.
The IDEaS problem is real and specific. Lombardi named it without softening: founders have won IDEaS challenges, celebrated the grant, and then discovered two things. First, there is no reliable pathway from winning a challenge to an actual procurement contract. Second, in a number of cases, winning an IDEaS grant effectively bars the company from bidding on related future work – because the government classifies it as a conflict of interest. The correct response to learning this, as someone in the audience noted, is laughter. But it is expensive laughter for the companies it happens to. The broader point is that government innovation programs, however well-intentioned, are not a substitute for commercial market discipline. Founders who build their strategy around grant funding without a parallel path to real customers are building on unstable ground.
The domestic market will not scale your business. Lombardi was direct on this too: Canadian government buyers are valuable early customers, but they will never buy at the volume required to build a serious company. The Five Eyes and NATO allied markets – the US, UK, Australia, and European partners – are where the real volume lives. Canadian defence companies should treat the domestic market as the validation stage, not the growth stage. Companies that internalize this early build products and supply chains capable of meeting allied procurement standards from the start, rather than retrofitting compliance after the fact.
Secrecy is killing companies that do not know it yet. Galloway made an observation that should be uncomfortable for a significant portion of the companies in the room. Founders who treat their product as too confidential to put into the hands of end users – who insist on NDAs before showing a design, who believe their IP is so sensitive it cannot be tested in the field – are consistently the ones who fail. The companies that win, by contrast, are the ones that find a way to get hardware in front of real operators under real conditions as early as possible, even when that is hard. He cited Sentinel R&D, which got on a plane to Estonia early in the Ukraine conflict and put their drones directly into the hands of warfighters when it was still difficult to do so. They got validation. They iterated. They built credibility that no trade show demonstration can replicate. The lesson is not that IP does not matter. It is that an untested product with tight secrecy is worth less than a field-validated product with a known iteration record.
Hardware has a strategic window that software does not. Galloway pushed back on the standard VC narrative that hardware is a harder investment than software. His point: in an AI environment where software moats are eroding faster than ever, hardware companies still have genuine technical differentiation that is difficult to replicate quickly. That window will not last forever, but it exists now – and it is one reason why defence hardware companies, specifically, are attracting more serious institutional capital than they were three years ago.
The communication gap is the deepest structural problem. Langlais identified what he called the core constraint in Canadian defence readiness: operational problems
evolve in weeks, but procurement and development cycles run in years. Getting feedback from the operator on the ground all the way up through command structures to a procurement decision and back down to a production change is a process measured in years, not weeks. Until that loop shortens, the only companies that will succeed in Canadian defence robotics are the ones that find informal pathways to operator feedback – that go directly to the end user, in the field, before they have a contract, before they have permission, before it is easy. The MINERVA Initiative– launched by the Canadian Army in December 2025 to accelerate general purpose uncrewed systems at the tactical level – is a deliberate attempt to shorten exactly this loop, by embedding soldiers as co-developers alongside industry from the start. It is worth watching closely.
The investors were not pessimistic about Canada’s trajectory. Galloway noted that Garage Capital has been investing in defence companies since 2014 and sees Canada as genuinely positioned to be a technology vendor to the world in this sector. But the message underneath the optimism was consistent: the opportunity is real, the barriers are structural, and companies that wait for the system to clear a path for them will be waiting a long time.
What This Means for Hardware Companies Entering the Defence and Dual-Use Market
The picture that emerged across ROAR is not discouraging. Canada’s defence spending trajectory – moving from roughly 1.5% of GDP toward 3.5% over the next several years – represents a structural increase in available capital. The dual-use market across public safety, critical infrastructure, natural resources, and defence is real and growing. The regulatory environment is evolving. The talent is here.
New ecosystem infrastructure is taking shape alongside the spending increases. BOREALIS– the Bureau of Research, Engineering and Advanced Leadership in Innovation and Science – is now active as Canada’s multi-departmental defence
innovation coordination mechanism, with its Defence Innovation Secure Hubs (DISH) program offering up to $50 million to create secure, mission-focused environments where industry, academia, and government can co-develop and test uncrewed systems and quantum technologies. These are not small signals. They represent a structural shift in how Canada intends to connect its innovation ecosystem to operational capability – and they create real entry points for companies that are ready to engage seriously.
But the structural risk in this market is the same structural risk in any complex hardware sector: the gap between a working prototype and a manufactured, fielded, maintained, iterated product is far larger than it appears at the demonstration stage.
The companies that will build durable positions in Canadian defence robotics – in UAV and UGV platforms, in counter-UAS systems, in drone-as-first-responder programs, in ISR payloads, in ground logistics robots – are the ones that treat manufacturing readiness, supply chain resilience, and operational robustness as first-order design requirements rather than downstream considerations.
That means selecting materials and architectures that survive field conditions, not just test ranges. It means designing for component-level repairability and modularity rather than platform-level replacement. It means building supply chains with redundancy from
the start, not after the first disruption. It means understanding that the thermal, vibration, ingress, and communications requirements of a system deployed in Canada’s Arctic are not the same as a system deployed indoors or in moderate climates – and that those differences need to be addressed in the engineering, not managed around in the field.
And it means being honest about the difference between a prototype that demonstrates capability and a product that survives volume production, operator use, environmental stress, and continuous iteration.
What Comes Next
The ROAR Conference series continues with its next event on November 18-19, 2026 in Ottawa at the Ottawa Convention Centre, focused on Defence and Government. If you are building or procuring unmanned systems for the Canadian government or defence market, it is the right room to be in.
Where Inertia Fits
Canada’s defence ecosystem has no shortage of innovative technology. What it lacks are organizations that can turn that technology into something deployable, manufacturable, and scalable.
Inertia occupies that gap. We are a Defence Capability Acceleration Partner – a Toronto-based engineering and manufacturing firm with 20 years of experience moving complex hardware from concept to field-ready product. We work across the full stack: requirements, system architecture, ruggedized design, embedded software, sensor integration, manufacturing readiness, and production. Not as a consulting firm. Not as a contract manufacturer. As an integrated execution partner.
What sets us apart is how we work. Inertia was built on venture-backed product development – clients with real capital pressure, hard deadlines, and no tolerance for scope drift. That operating rhythm is embedded in how we run every engagement. We move with the urgency of a startup and the engineering rigour of a prime, without the overhead of either.
Where most organizations can prototype, Inertia can industrialize. Where most defence innovators stall between a promising TRL and a deployable system, Inertia closes that gap – in Canada, at pace.
The Canadian defence and dual-use robotics sector is moving from the demonstration phase into the production phase. The companies that win the next five years will not be the ones with the most impressive concept vehicles. They will be the ones that can produce reliably, iterate quickly, and build hardware that holds together under real operational conditions.
If that is the transition you are navigating, we would be glad to talk.
