Dr. Chaouki Kasmi is the President of Technology and Innovation at EDGE Group.
getty
There’s a (likely apocryphal) story about a drone company in Ukraine that, as a signature promotional strategy, would send one individually wrapped cookie along with each of their drone shipments. These cookies were incredibly popular, a much-loved gimmick. One day, a soldier unpacked a shipment, tasted the cookie and realized it was nothing special—a little dry, sugary, crumbly. Nothing to write home about.
“Why is everyone so excited about these cookies?” he asked his commanding officer.
“It’s not the cookies; it’s the shelf life,” his commanding officer told him. “If the cookie is fresh, the drones are good. If the cookie is stale, we know the drones are already obsolete.”
This story illustrates one of the axioms of technology development today. For years, the focus has been on pure innovation, usually around software; governments have been willing to absorb the high costs of having the most sophisticated systems possible.
But recently, a new reality has been exposed—perhaps nowhere more starkly than in the Russia-Ukraine war: Low-cost, rapidly produced, iterative systems (especially drones) are changing the game of warfare, showing that outsized damage can be inflicted on slow, expensive platforms, no matter how sophisticated their design. This has created a fundamental structural imbalance: Defenders must spend massively to counter threats from systems that are cheap to make, tweak and deploy.
As a result, technology—in defense, certainly, but also in other industries such as automotive, aerospace, energy, even biotech—can no longer optimize only for performance. Instead, they must design with manufacturing and cost top-of-mind from the start.
Learnings From The Automotive Industry
Interestingly, the auto industry has already figured out what defense industry executives are (rather expensively) learning right now; they’ve been much quicker to internalize that every component’s cost at scale has a direct impact on margins. Consequently, they’ve been designing cars for years with the understanding that manufacturability and cost-efficiency are the primary constraints, and they compete internationally to make vehicles not just advanced but also cost-effective.
The defense industry has, at least until now, often behaved the opposite way. Because executives were secure in the knowledge that the government clients were prepared to pay for sophistication, price was a secondary concern. But now, the whole sector’s hand is being forced by the paradigm we see in wars overseas: Optimizing designs necessitates making them easier and cheaper to produce at scale, as opposed to merely creating the highest-performing version.
Designing For Manufacturing
While a return to manufacturing focus might sound, on the surface, like a step back from the proliferation of AI-driven technology, in fact the opposite is true.
Previous generations used statistical techniques to improve manufacturing performance. Today, AI carries this load and more, compacting the matrix of performance, cost, manufacturability and supply chain risk into a single optimization problem. This enables faster, better, more cost-aware design.
At the technical level, AI connects design tools, engineering models, supply-chain data and production line telemetry. This allows design for manufacturing and design-at-cost to be embedded in the processes from day one. Production lines and machines all become part of an agentic AI-enabled framework that constantly seeks to maximize throughput and reduce human workload.
At the workforce level, roles have evolved: Traditional technicians are now replaced by AI operators overseeing and refining the AI-enabled workflows; in other words, your manufacturing workers of the past are today’s humans in the loop. This poses undeniable (but necessary) challenges, not only as far as the tools that must be adopted but also in rewiring organizations’ processes, data flows and responsibilities around these new AI tools.
AI-Powered Resilience
With the preeminence of low-cost, fast-production items and the increased reliance on AI, manufacturing strategy is today’s strategic differentiator.
Vertical integration (that is, when a company takes control of multiple stages of its own supply chain, from raw materials to production to retail and distribution, rather than relying on a third party) has become increasingly important in an environment where profit and competitiveness now come from scale, as opposed to a small quantity of high-margin units. Owning or tightly orchestrating production capacity enables faster ramp-up, adaptation and iteration.
To be clear, “in-house” production does not mean everything occurs on a single campus or even within one corporation. It may involve template production lines replicated across geographies with tightly aligned partners. The key is a strong link between design and production, with enough shared information to move production between sites efficiently.
AI can help further vertical integration by mitigating issues such as supply chain fragility (say, semiconductor shortages, or situations wherein certain important components can suddenly become unavailable, end-of-life or monopolized by a single large order). By continuously monitoring stocks, lead times, geopolitical factors and component life cycles, AI can flag when parts are no longer viable or at risk of scarcity and suggest alternative components or subsystems. This type of monitoring enables R&D and production to continue with fewer delays, even when supply chain breakdowns are otherwise creating bottlenecks across an industry.
Competing As A Non-Manufacturer
Over the years, many companies and even entire regions have outsourced or shut down their industrial capacity, largely for cost-saving reasons; simply, it was cheaper to produce products elsewhere. But today, there is a clear trend to rebuild that capacity. When they don’t control their own manufacturing, nations and firms are strategically exposed.
For organizational leaders, the prospect of starting up manufacturing capacity from scratch can appear daunting. But paradoxically, starting from scratch can create an advantage. New production lines can be built natively automated from day one, with AI integration and other efficiencies.
Software-only companies, however, do face a structural vulnerability. AI has made it easier and faster to replicate applications, and once the idea and user experience are known, the software itself often ceases to be a lasting moat. Software companies can find a durable advantage, but that increasingly requires combining software with tailored hardware, building cost-efficiency at the forefront and laying the foundation to support a continuous update and production cycle.
Bridging Prototype And Production
Ideation is no longer today’s critical bottleneck. Instead, it’s the transition from prototype to industrial-scale production.
In traditional production models, first you built prototypes, then you waited for full requirement validation. Only then would you even think about industrialization.
But companies moving at that pace today will lose ground. The emerging model is different: The entire journey—prototyping, testing, industrialization, deployment—must be treated as a single, connected process. Only this way can time-sucking rework be minimized, as early designs are already encoded with AI-determined manufacturing constraints and supply chain realities.
The companies at the forefront of today’s manufacturing boom will be able to develop, test and produce quickly. They’ll maintain an end-to-end digital thread from the first design to the factory floor, enabling them to respond quickly to new requirements and constraints on both time and cost.
The future belongs to the next generation of tech leaders who can further collapse the distance between innovation and production, leveraging AI and manufacturing capacity together seamlessly to deliver at scale. And if their product comes with cookies, well, so much the better.
Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?
>> Home