Annons

When CAD Meets the Shop Floor Reality: The Hidden Design Flaws Fueling Multi-Million Dollar Waste

Encube Innovator HUGO NORDELL’s Promise: “Halving time-to-market and slashing production costs by 30% as AI enters the shop floor.”
Does that sound like an unrealistic stretch of operational marketing prose? It certainly could be, but it doesn’t have to be. With AI now a tangible part of the equation, what seemed impossible just a few years ago is fast becoming reality.
“That is exactly the case,” says Hugo Nordell, Cofounder and CEO of the Swedish deep-tech venture Encube. “And yes, our solution can cut development cycles in half and lower production costs by up to 30 percent. This is demonstrated by data from Encube's pilot projects with large industrial enterprises. The secret lies in a native AI platform, architected from day one to power a continuous, automated design review process across the entire R&D pipeline. By establishing a synchronized engineering collaboration engine all the way down to the factory floor, Encube accelerates workflows, ensures higher data accuracy, and ultimately paves the way for significantly more profitable hardware development, from concept to production.”
But looking closer under the hood: what exactly is it that Encube addresses?
Within the global manufacturing landscape, there is a recurring and notorious disconnect between the drawing board and what actually takes place on the shop floor. A classic scenario almost always involves a design that looks digitally flawless within a CAD environment, yet carries one or several hidden vulnerabilities. It might clash with the laws of physics, suffer from flawed or inadequate material properties, or simply demand a manufacturing capacity that isn't available in real-world production. In physical reality, this kind of misalignment can carry an immense price tag—sometimes running into the billions. This is precisely what Hugo Nordell believes can be avoided with the Encube platform, where a key objective is to shift risk analysis directly into the CAD stage.
PLM&ERP News’ Verdi Ogewell caught up with Nordell at the company's headquarters in Stockholm, Sweden. He brings a remarkably seasoned background as an industrial innovator, with heavy, production-oriented executive roles behind him—including leading digital excellence and growth ventures at Swedish industrial giant Sandvik, as well as serving as Senior VP for the Norwegian industrial conglomerate Aker Solutions. Furthermore, his track record includes strategic founder and CTO roles in San Francisco, alongside Silicon Valley projects spanning advanced technical domains such as computer vision, machine learning, and connected drone technology. In short, Nordell knows precisely what he is talking about when it comes to industrial production.
Backed by a total of $23 million (SEK 215 million) Series A funding, Encube is now accelerating its commercial rollout, supported by a heavy-hitting international investor trio consisting of Kinnevik, Inventure, and space-focused Promus Ventures. This substantial capital injection is set to scale the company's cloud- and AI-based platform for hardware and industrial design through an aggressive expansion in Europe and strategic industrial partnerships. That the proprietary technology delivers world-class performance is evidenced by an already strong market validation with participation of major players like Volvo Group, Scania, and Beyond Gravity.

Plagued by a Structural Catch-22
The hardware sector has long been plagued by a structural Catch-22: while the software industry has sprinted forward at breakneck speed over the past decade, heavy industry has remained trapped in a fundamentally different, uncompromising reality. Demands for operational productivity are skyrocketing, yet tooling and manufacturing engineers are routinely forced to rely on software architectures still deeply rooted in the late 1990s. This underscores one of the manufacturing world’s most frustrating, costly secrets: the staggering capital literally thrown into the wastebasket every single year during the industrialization phase. Resolving this deep-seated friction requires a radical shift away from engineering guesswork and toward absolute deterministic certainty—precisely the structural paradigm that Nordell and his team are disrupting.

TRADITIONAL HARDWARE DEVELOPMENT has long been confined to heavy, localized desktop workstations, creating a rigid barrier between core engineering and the wider enterprise.
To break this silo, Encube is transforming the web browser into an agile workspace where complex product data and cross-functional teams finally coexist. Hugo Nordell, co-founder and CEO of Encube, describes the platform as a “massive digital whiteboard experience.” “The platform allows users to seamlessly integrate any graphical asset, transforming it instantly into an interactive 3D asset right alongside active engineering spreadsheets,” Nordell notes. This workspace enables distributed teams to annotate designs, capture feedback, and track revisions in real time. Crucially, this browser-based architecture democratizes technical data by removing the historical dependency on high-end local hardware. “Local computing power is no longer a barrier to entry; the system is built to ensure seamless accessibility for everyone, whether they operate in procurement or deep mechanical engineering,” Nordell emphasizes. By enabling multiple stakeholders to view, analyze, and comment on the same design simultaneously, the platform turns fragmented engineering discussions into actionable, real-time steps. It is a vital shift toward moving risk assessment earlier in the lifecycle, allowing the entire enterprise to collaborate at the speed of thought.

“Our platform is entirely focused on intelligent predictability,” Nordell explains. “By leveraging domain-specific AI to identify hidden design and manufacturing risks long before they ever reach the factory floor, Encube eliminates the costly physical errors that have historically drained industrial margins.”
When a new product is drafted within a modern CAD environment, it is incredibly easy to be lulled into a false sense of mathematical perfection. However, the underlying methods and core digital frameworks powering these tools have, in many cases, completely failed to keep pace with modern industrial demands.
The chasm between a digital solid model and the unforgiving physical reality of a multi-axis CNC machine, a heavy press tool, or an automated welding line remains monumental. Yet, it is at the absolute inception of this design process that the most critical, cost-defining decisions are made—all too often completely in the blind. Without timely, integrated Design for Manufacturing (DFM) analysis, microscopic tolerance deviations, flawed wall thicknesses, or unfavorable material choices pass through undetected, later rendering a component either entirely impossible or prohibitively expensive to produce. A minor, misplaced bracket or a suboptimal radius within the CAD environment quickly translates into catastrophic micro-cracks under pressing or severe thermal stress during physical assembly.

WHEN A NEW HARDWARE DESIGN IS LOCKED IN, most of its downstream cost and manufacturing complexity has already been decided. Yet, the teams making those early, critical choices are all too often operating in an information vacuum. To bridge this gap, Encube’s AI engine analyzes geometric data in real time to evaluate manufacturability long before a design ever reaches the shop floor. By flagging potential structural bottlenecks early, the platform shifts risk management directly into the design phase—preventing the costly errors that historically surface only during physical production. “Our architecture is engineered to offload the repetitive yet critical analytical burdens that engineering teams encounter every day,” notes Hugo Nordell, co-founder and CEO of Encube. Nordell, whose background spans theoretical physics and leading global software divisions at Sandvik, witnessed these exact friction points across the manufacturing sector, particularly in complex CNC machining. “Regardless of the enterprise, early-stage design geometry dictates the downstream financial reality. However, traditional workflows fail to provide cross-functional teams with the necessary insights at the moment of inception,” Nordell explains.

Collaboration, Alignment and Key Decisions Outside Existing Software Suites
To understand the core of the problem, one must examine the classic, step-by-step engineering design flow. Looking at the actual software landscape reveals a glaring, structural paradox. On one side, engineers utilize powerful design and simulation tools like Siemens NX/Designcenter, SolidWorks, and Ansys to shape the product’s geometry. On the other side, massive enterprise PLM platforms like Teamcenter, 3DEXPERIENCE, Windchill, and SAP manage lifecycle data and corporate processes.
Sandwiched between them lies the manufacturing software industry’s most frustrating bottleneck: a chaotic zone where 60 percent of the actual day-to-day work—collaboration, alignment, and critical decision-making—takes place.
This middle ground is defined by what can only be described as ”integration spaghetti.” As illustrated in the architectural breakdown, a chaotic web of custom lines, fragmented screenshots, emails, and manual translations is constantly required just to move engineering data from the CAD design environments on the left to the enterprise PLM networks on the right. Because the traditional suites fail to provide a unified, intelligent bridge, the most vital cross-functional decisions are effectively forced outside the primary software infrastructure, draining both time and data integrity.

The integration spaghetti. The chaotic web of custom lines and manual translations required to move engineering data from CAD design environments (left) to enterprise PLM networks (right).

This intricate tangle of curved lines represents the very real friction of industrial engineering. Moving data seamlessly between the digital design canvas and the enterprise backbone remains a monumental chasm—a highly fragmented space dominated by brittle integration scripts, manual file conversions, and broken feedback loops.
Without timely Design for Manufacturing (DFM) insights embedded directly into this active pipeline, critical flaws inevitably pass through unnoticed. A microscopic tolerance deviation or material mismatch defined early in the CAD environment gets lost in the spaghetti of data transfers, only to resurface weeks later when the physical production line grinds to a halt.
Crucially, hardware carries a fundamentally different risk profile than software. On a digital screen, an integration glitch is merely a bad data sync; in the physical world, an undetected design failure in a brake drum or a robotic arm can cost human lives. Resolving this profound systemic vulnerability requires moving entirely away from engineering guesswork and toward absolute deterministic certainty.
This is precisely the frontier Nordell and his team are transforming. Encube’s core objective is to completely streamline this chaotic crossroads. Rather than weaving more custom spaghetti lines into an already fractured web, their domain-specific AI acts as an intelligent, unified layer that bridges the design silo and the PLM backbone. By introducing intelligent predictability right at the volatile intersection of design and data management, the platform effectively neutralizes manufacturing flaws before they are ever locked into the enterprise ledger.

PIONEERING A PARADIGM SHIFT IN HEAVY INDUSTRY: Encube co-founders Johnny Bigert (left) and CEO Hugo Nordell (right). Bringing together a powerful combination of disciplines, the founding duo is spearheading a structural transformation in how heavy industry approaches product realization. Bigert, a seasoned Skype and Klarna veteran, brings deep expertise in building frictionless, hyper-scalable software ecosystems. Nordell, trained as a theoretical physicist with a background in Silicon Valley technical development and executive roles at industrial giants like Sandvik and Aker Solutions, masterminds the platform’s deep engineering-grade logic. By anchoring their browser-based platform on a proven geometric kernel and domain-specific, physics-informed AI, Bigert and Nordell are effectively erasing the traditional, costly chasm between the digital blueprint and shop-floor reality. Their shared vision tackles one of manufacturing’s greatest modern threats: the loss of irreplaceable manufacturing knowledge amid accelerating retirement waves, at a time when Europe feverishly attempts to rebuild its regional industrial competitiveness. Through Encube, they have built a unified environment that democratizes data and allows distributed enterprise teams to neutralize risks, resolve engineering friction, and collaborate seamlessly at the speed of thought.

Precisely Where the Shoe Pinches
”That is precisely where the shoe pinches,” explains Hugo Nordell, as we discuss the manufacturing sector’s inherent conservatism and the role AI must play to counter it. “Companies simply have zero room for error. This high stakes reality causes the perceived implementation risk for new core systems to skyrocket, which in turn locks the industry into its legacy tracks.”
Nordell brings a seasoned perspective forged at the highest levels. After years of developing cutting-edge robotics and autonomous technology in Silicon Valley, he returned to assume pivotal executive roles within European heavy industry. Across both environments, he witnessed the exact same destructive patterns repeating themselves, irrespective of whether the enterprise was a niche specialist or a global industrial giant: design flaws were almost universally discovered only when the physical machinery began running on the factory floor.
“It is not necessarily the production execution itself that fails,” Nordell notes, shaking his head slightly. “The fundamental flaw lies in the profound gap between the design’s original intent and the raw, unforgiving reality required when the product is actually set to be manufactured.”
This systemic chasm has become painfully acute in light of the tectonic macroeconomic shifts of recent years. The era of ultra-low inflation, cheap energy, and frictionless globalization that propelled the industry forward for two decades is definitively over. Today, European manufacturing companies are forced to navigate a volatile landscape defined by geopolitical protectionism, fluctuating commodity prices, and a sharp interest rate shock that has dramatically driven up the cost of capital.
In this new economic reality, tying up millions of dollars in protracted R&D cycles and capital-intensive tooling modifications is no longer just operational inefficiency—it represents a direct, unsustainable financial risk. Compounding the pressure, the industry is simultaneously squeezed by an acute shortage of senior engineering expertise, right at the peak of a massive wave of nearshoring, where production is feverishly relocated closer to home markets to secure vulnerable supply chains.

RESTORING STRATEGIC CONTROL IN PRODUCT REALIZATION. An engineer evaluating structural assembly details via a web-based 3D workspace. In legacy hardware workflows, design evaluations and manufacturing risks are traditionally isolated in a post-modeling silo, leaving cross-functional teams completely blind to downstream financial and operational consequences. This structural latency leaves enterprises with narrow, unpalatable choices late in the development cycle: either absorbing severe margin erosions or accepting crushing time-to-market delays. By integrating deterministic, physics-grounded AI directly into the live modeling interface, this volatile crossroads is completely streamlined. The platform dynamically flags latent geometric vulnerabilities and manufacturing complexity, shifting critical risk management straight to the moment of inception. ”Companies are constantly being forced into operational compromises that nobody actually wants,” observes Encube co-founder and CEO Hugo Nordell. ”But the earlier we can expose these hidden engineering risks, the more strategic headroom engineers gain to maintain both tight schedules and budgets. It is fundamentally about restoring control to the team.”

Forced into Compromises
When dispersed specialist teams, global suppliers, and isolated software tools are forced to collaborate within this harsh economic climate, coordination rapidly devolves into a logistical nightmare. This is precisely where bottom-line profitability quietly erodes. By the time critical issues are finally detected, enterprises are left to choose between two equally unpalatable alternatives: launching a more expensive product with severely degraded margins, or returning to the drawing board at the cost of devastating delays.
“Companies are constantly being forced into operational compromises that nobody actually wants,” says Hugo Nordell. “But the earlier we can expose these hidden engineering risks, the more strategic headroom engineers gain to maintain both tight schedules and budgets. It is fundamentally about restoring control to the team.”
For heavy industrial players like Yaskawa Nordic AB—whose advanced automation and robotics solutions demand absolute precision at the component level—this insight is mission-critical. This is precisely where Encube’s platform steps in: not to replace the human element, but to serve as a sophisticated digital copilot explicitly calibrated for industrial complexity.
Yet, Nordell remains fully aware of the deep-seated skepticism that permeates the engineering market.
“The industry’s demands for reliability are absolute, and that has historically made it incredibly difficult for new software tools to establish a meaningful foothold,” Nordell notes. “However, despite the widespread and understandable AI skepticism, recent technological advancements have unlocked entirely new capabilities to build systems that truly comprehend complex hardware development. But let’s be clear: generic chat functionalities simply won’t cut it here.”

DEMOCRATIZING COMPLEX PRODUCT DATA: Encube’s browser-based platform acts as a unified digital workspace where interactive 3D CAD geometry, engineering spreadsheets, and operational project data finally coexist. Rather than forcing distributed teams to rely on fragmented screenshots, disconnected emails, or localized desktop workstations, this shared environment allows mechanical engineers, procurement experts, and manufacturing suppliers to review assemblies and analyze potential design flaws simultaneously. By grounding its agentic AI workflows on a proven Siemens Parasolid geometric foundation rather than simplified approximations, Encube provides teams with real-world, engineering-grade feedback on manufacturing complexity and cost drivers in real time. This intuitive, ”digital whiteboard” architecture shifts risk assessment to the absolute inception of the product realization process—enabling the entire enterprise to neutralize engineering vulnerabilities and collaborate seamlessly at the speed of thought.

Domain-Specific Models
To overcome these baseline limitations, Encube is training its own domain-specific AI models designed to directly factor in physical manufacturing constraints. By consolidating cross-functional data into a singular, transparent digital environment, traditional organizational silos are effectively dismantled, democratizing deep engineering expertise across the enterprise. This places AI-driven insights into an actionable, real-world context, making advanced design evaluations instantly accessible in real time.
The financial and operational impacts of this shift redefine the traditional boundaries of hardware development. When deployed across active development programs, empirical data reveals a fundamental transformation: time-to-market is slashed by up to half, overall production costs are reduced by 20 to 30 percent, and engineering productivity is effectively doubled.
Yet, from a broader, macroeconomic perspective, this paradigm shift is about far more than merely polishing the balance sheets of individual enterprises; it is a matter of industrial survival on a global stage. While the United States has pioneered the software revolution through Silicon Valley, and China has constructed a near-unstoppable infrastructure for large-scale, heavily subsidized mass production, Europe’s primary competitive edge has historically resided in its deep, highly refined engineering expertise. However, that sovereign advantage is rapidly eroding as Western demographics shift, and accelerating retirement waves systematically drain factories of invaluable, tribal experience.

European Industry Must Compete on Pure, Uncompromising Efficiency
If European industry is to hold its ground against American tech hegemony and Asian manufacturing giants, it cannot afford to compete on cheap labor or massive state subsidies. Instead, it must compete on pure, uncompromising efficiency.
Hardware-focused AI like Encube could very well be the missing puzzle piece. By institutionalizing generations of tribal engineering knowledge directly into a deterministic algorithm, Europe can leapfrog costly, physical iteration loops to build a new ecosystem—one where the pace of hardware development finally matches the velocity of software, without ever compromising on quality.
Ultimately, Encube’s emergence demonstrates that AI’s next major breakthrough is not happening in generic chatbots, but at the very heart of heavy industry. In a highly volatile global economy, the ability to close the persistent chasm between the digital blueprint and the factory floor represents the definitive difference between a triumphant flagship project and a costly industrial wreck.

The Next Frontier: Redefining Product Realization
”What about the future? The ideas behind what Encube puts on the table has a lot to do with CAD and what is happening in this domain,” I asked Nordell. A defining edge of Nordell’s vision lies in his view of the future, where AI is a significantly heavier and more dominant element across the entire product realization process than it is today. In what way?
The deeper one scratches the surface of today’s CAD landscape, the clearer it becomes that traditional PLM giants have mostly built sophisticated digital drawing pens focused on localized drafting speed. Nordell and Encube, by contrast, look to transform software into a proactive physical and commercial engine operating far beyond the engineering department.
This requires shifting the entire center of gravity in product realization.
“Hardware development is a balancing act between aesthetics, functionality, and production economics,” Nordell observes. “Yet today, costly misalignments between engineering intent and the strict constraints of the factory floor are uncovered far too late. By embedding physical laws and manufacturing rules directly into a browser-based environment, Encube allows the entire enterprise to witness how a single design change impacts both the budget and the physical production lines in real time.”

FEM and  the VALUE OF SHIFTING LEFT. In hardware product realization, there is an invisible, high-stakes boundary line between geometric form and physical reality. You can sketch a mathematically perfect model in CAD, but real-world physics doesn’t care about a digital drawing pen. That is why engineering teams rely on Finite Element Analysis (FEA) to break complex structures down into smaller ”elements” and predict structural behavior before cutting steel. FEA has historically been a disconnected, post-design validation step. By the time a simulation reveals a structural flaw or material weakness, the design is already locked in, forcing teams back to the drawing board. ”The core objective is to activate these risk assessment measures much earlier in the product realization process,” explains Encube innovator Hugo Nordell in today’s article. This is the true power of a ’shift left’ strategy in engineering. By embedding physics-informed AI and deterministic logic directly into the active modeling pipeline, the next generation of industrial software absorbs the solver mechanics entirely. Instead of waiting weeks for traditional stiffness matrices to crunch, engineers get instant feedback on real-world stress and material performance during the design phase. Ultimately, the future lies in dissolving the traditional CAD/CAE chasm. It is about deploying physics-grounded intelligence so enterprises can neutralize engineering risks from day one.

Shifting the Paradigm: From Form-Finding to First Principles
Looking ahead, this paradigm shift opens up a highly compelling, speculative frontier for both generative CAD and downstream engineering simulation. While the market is currently flooded with flashy social media demonstrations of prompt-to-geometry generation, Nordell maintains that true generative CAD must evolve beyond superficial form-finding. The future lies in AI architectures that don’t just output random shapes, but instead synthesize complex engineering logic from first principles.
Crucially, this architectural shift extends far beyond geometry; it inherently redefines how we approach Finite Element Method (FEM) solvers. In legacy workflows, structural integrity and advanced material properties—such as stress tolerance, thermal conductivity, and elasticity—are evaluated long after the geometric form is locked in. By unifying CAD and CAE into a singular, continuous feedback loop, engineers can instantly see how alternative materials or subtle structural adjustments perform under real-world physical stress, eliminating months of costly, iterative guesswork.

IN MODERN PRODUCT DEVELOPMENT, the challenge is no longer just about creating innovative products—it is about ensuring that global R&D, supplier, and manufacturing teams can review and improve complex product data efficiently. Yet today, the industrial ecosystem remains bottlenecked by disconnected design reviews, fragmented screenshots, and desktop-only CAD access that slow decision-making.
To break this barrier, Encube is redefining cross-functional collaboration by bringing all stakeholders into a shared, browser-based 3D environment. At the heart of this engine lies Siemens Parasolid, the market-leading geometric modeling kernel. By integrating Parasolid as its core foundation, Encube ensures that its multiplayer workflows and agentic AI are grounded in precise, engineering-grade geometry rather than simplified approximations. This allows the platform’s AI to reason about design intent with uncompromised fidelity. Ultimately, the goal is collaboration ”at the speed of thought.” By combining agentic insights with high-fidelity browser visualization, Encube democratizes deep technical data across the enterprise, allowing distributed teams to neutralize risks and take action earlier in the lifecycle.

Absorbing the Mechanics of the Solver
Rather than treating finite element analysis as a disconnected, post-design validation step that requires separate, heavy computing environments, the next generation of physics-informed AI models will essentially absorb the solver mechanics. Instead of an engineer manually setting up a boundary value problem and waiting for traditional iterative or direct linear solvers to crunch a stiffness matrix, AI neural operators will instantly predict stress concentrations, thermal gradients, and structural displacement during the active modeling process.
This holistic philosophy poses a profound challenge to legacy players like Autodesk, Dassault Systèmes, Siemens, and PTC. While these giants use large language models and predictive assistants to automate 2D drawings or speed up isolated geometric modeling, they largely operate within an isolated engineering bubble.

Breaking Ranks with Deterministic AI
Encube breaks ranks by pioneering deterministic, physics-based AI. It is an approach that carries significant industry validation, validated by the arrival of Mats Friberg as Encube’s newly appointed Senior Advisor. As the well-known, recently retired head of Siemens Digital Industries Software’s Nordic operations, Friberg’s insider perspective highlights just how fast the new ecosystem is moving:
“It is easy to recognize the inherent strength of the Encube concept and the profound cost savings it can unlock,” Friberg shares. “Consequently, it comes as no surprise that the company has already secured partnership agreements with key players like Siemens—both spanning their core Parasolid kernel technology and in developing deep integrations with Designcenter/NX CAD and Teamcenter PLM. Furthermore, Encube has successfully developed an integration with SOLIDWORKS, and we are actively engaged in ongoing dialogues regarding similar initiatives with both Dassault Systèmes and PTC.”
“This is enabled by physics-informed and generative AI that combines empirically proven physics with symbolic reasoning to deliver grounded, reproducible outputs,” Nordell explains. Ultimately, the future of product realization lies not in making humans draw faster, but in deploying shared intelligence to help entire enterprises make vastly smarter decisions from day one.

Print Friendly, PDF & Email

Success Stories

Industriellt

Success Stories

Intressant på PLM TV News

Aktuell ANALYS

Aktuell Analys

Aktuell Analys

3D-printing

Block title