It’s 2026 and Europe is running out of time for experiments. Fortunately, the technology is ready: European deep tech is moving to deployment. The need is urgent, the science works, and we can now scale from promise to meaningful impact in Europe and beyond.
Based on our insights as an investment team, one focus point for 2026 is clear: we’re entering the year of the data dividend.
From Promise to Deployment
Europe has reached a turning point. The deep tech ecosystem is maturing at the same moment the continent faces pressures that demand real, working solutions.
Consider all these challenges: Manufacturing is under strain from renewed global competition and shifting trade terms. Defence procurement has moved from long planning cycles to urgent delivery. Europe needs secure access to space and a strong homegrown space industry, including independent access to orbit and space-based intelligence that underpins modern power. Grid instability is pushing industry to demand steady, reliable energy as data centres draw ever more load. At the same time, the care crisis has arrived. The working-age population is shrinking while the elderly population grows, making robotics a necessity for society, not a nice-to-have for productivity.
These are not distant challenges. They require deployable solutions, soon. Fortunately, across each of these fronts, deep tech is beginning to deliver, built on scientific foundations that are no longer theoretical but usable. Hardware progress continues, though it still takes time to become robust and affordable. In parallel, software and AI advances now allow rapid deployment on existing systems, turning years of research into tools that work in the real world.
Within this shift, one issue stands out: data.
Whether AI succeeds will depend less on model size and more on how growing volumes of data are captured, structured, and used. Data alone creates no value. Value comes from turning it into insight and action. The leaders will not be those with the most data, but those who build the systems that convert physical and biological reality into usable knowledge. Much as Lean Production once reshaped manufacturing through discipline and structure, these new infrastructure layers turn wasted data into lasting advantage.
This is what we at Vsquared call the Data Dividend: the compounding value created when materials science, computation, and systems engineering converge under real adoption pressure. Deep tech deployment accelerates not because the technology suddenly appears, but because urgency forces use just as the pieces fall into place. For years, capital was the missing link. As it now begins to flow, data potential is turning into measurable impact. We expect this to define 2026.
The Data Dividend: Lean Production 2.0
Across deep tech verticals, we are witnessing a structural shift. Industries are moving from an era where companies compensated for poor-quality data by accumulating vast quantities of it, to one where data is precise, well-organised, and immediately actionable. As a result, the focus is no longer on volume, but on data that preserves causal signals, temporal dynamics, and biological heterogeneity rather than averaging them away.
Lean Production offers a useful analogy here. It once created exponential productivity gains through process innovation, turning waste elimination and continuous improvement into compounding advantages. The Data Dividend represents Lean Production 2.0 for the current era. Where Lean optimised physical workflows, the Data Dividend optimises information flows.
This creates real economic value for the entire business world. Thanks to the data dividend, a single contextualised observation can prevent a million-euro production line failure, accelerate drug discovery by months, or enable predictive maintenance that doubles equipment lifetime. The ‘dividend’ per se derives from the delta between the cost of instrumentation and the exponentially growing value over time of preventing failures, speeding discoveries, and improving operations.
In other words, the Data Dividend is all about the compounding of sustained investment in sensors, automation, and computational tools. Like Lean Production’s kaizen, continuous improvement through incremental gains, each observation generates disproportionately higher value because it is instrumented, contextualised, and usable from the outset.
Interestingly, this value is inseparable from the physical systems that generate it. That’s because in deep tech, data-driven software and hardware co-evolve: models extracted from measurements are reliable only on the specific hardware systems they’re tuned for. A learning controller for one robot cannot simply transfer to another, it is specific to that system’s physics and constraints. From a company perspective, this hardware-software interdependence creates strong lock-in effects and high switching costs. Thus the Data Dividend compounds on two levels: generally improving data quality over time whilst simultaneously deepening system-specific advantages and business defensibility.
This dynamic is evident in our portfolio company Neura Robotics. In industrial automation, robot hardware is no longer the bottleneck. Deployment speed, system reliability and uptime are. But the real differentiation lies deeper—building partnerships where risk and upside are shared. This mirrors Lean Production’s emphasis on supplier partnerships and just-in-time coordination. At Neura, success (and therefore defensibility) comes from co-developing business models driven by urgency and competitive pressure, iterating together with customers over months until novel productive automation systems emerge. This human and organisational layer creates advantages that pure technology alone cannot.
The same pattern emerges across domains. In drug discovery, failed experiments become as valuable as successful ones when properly instrumented—the Lean Production principle of learning from every cycle applied to R&D. In energy management, real-time grid data enables software-defined power distribution—but only when paired with novel hardware infrastructure. Like switches in IT networks, power electronics, HVDC lines, and solid-state technologies form the physical substrate that makes software control possible. The Data Dividend in energy requires both the sensing layer and the actuation layer—the full stack from data to physical response. In defence systems, sensor fusion creates situational awareness from disparate inputs. The Data Dividend accrues to those who own the layer that transforms raw signals into deployable systems.
Two Forces Reshaping Every Sector
The Data Dividend manifests through two fundamental forces transforming every sector: the sustainability imperative, and the diffusion of intelligence across systems.
The Sustainability Crisis
Energy has become the foundational enabler for everything else. The shift from passive commodity to programmable resource accelerates as industrial players demand reliability at lower cost. Europe already has the world’s highest energy prices, threatening our economic competitiveness. Investment in new infrastructure is unavoidable, but it must deliver both resilience and cost reduction. Long-duration storage, hybrid generation systems, and software-defined grids create resilience whilst driving down system costs through optimisation. This is Lean Production applied to energy: eliminating waste, reducing variability, optimising flow.
This transformation propagates across every innovation layer. Higher energy density drives AI capabilities in smartphones and wearables. Electro-optics in data centres achieve higher data rates at lower power consumption. Energy is no longer another vertical, but the substrate enabling the entire Data Dividend.
The Diffusion of Intelligence
The second force is the systematic embedding of intelligence across all systems. As a result, strategic leverage shifts to the platforms closest to data generation—the systems that make raw signals AI-ready. These platforms form the quiet backbone of the new order, creating defensible value through technical difficulty and capital intensity rather than through consumer-facing applications:
Healthcare is shifting from reactive to preventive care. Consumer platforms like Neko Health, Oura and Apple Watch generate continuous longitudinal data streams. Meanwhile, startups like our portfolio company Quantune push wearable technology to the next level by measuring biomarkers such as glucose and lactate noninvasively, addressing what has long been considered a holy grail in health monitoring. These platforms instrument biology at consumer scale, making prevention economically superior to treatment. In parallel, patient-centric platforms make clinical data actionable for personalised medicine in reactive care settings. The key value driver across both domains is longitudinal data that tracks individual health over time, enabling lower-cost healthcare systems. This mirrors Lean Production’s ultimate logic: eliminating defects before they occur costs far less than fixing them later.
In robotics and manufacturing, record installation levels mask the real challenge: making robots programmable by non-specialists. This unlocks two transformations: shifting automation from high-volume to high-mix, low-volume tasks, and scaling automation directly at client sites without integrator dependency. This democratisation of robotics mirrors how Lean Production empowered factory workers to stop production lines and fix problems. The promise is data-driven robots that learn tasks from application experts. But the challenge remains formidable: getting AI-enabled robots to learn from non-experts whilst achieving industry-standard reliability. Current AI still struggles with this reliability requirement, making it the critical bottleneck between promise and deployment.
The demographic crisis amplifies the urgency. As labour supply contracts, robotics becomes a necessity. The value is not more advanced hardware, but the software layer that unlocks existing assets—maximising utilisation rather than increasing CapEx. Critically, this isn’t about individual robots doing isolated tasks, but complex automation systems where whole chains of actions and different systems connect through a software orchestration layer. Modern automation requires coordinating multiple robots, sensors, and systems into adaptive workflows. The software becomes the conductor, transforming disconnected hardware into coherent production symphonies.
Compute architectures are evolving in parallel. Memory-centric architectures that compute where data resides already demonstrate 10-100x reductions in AI energy consumption. But this is only one dimension. Photonics and electro-optics innovations are revolutionising system design. The shift from general-purpose GPUs to inference-specific systems reflects the industry’s move toward specialised, power-efficient architectures. Compute transformation now occurs at three levels: (1) memory-focused (e.g. in/near memory, novel memory types) architectures with inference-specific chips, (2) edge computing processing where data originates, and (3) electro-optics innovation spanning manufacturing to components.
In financial infrastructure, atomic swap infrastructure promises faster, cleaner settlement without trusted intermediaries. Privacy-preserving on-chain solutions are enabling fintech 3.0 companies to reach large revenues at unprecedented speed. This rebuild of financial plumbing enables programmatic settlement and automated compliance—applying Lean Production’s zero-defect philosophy to financial transactions.
Finally, at the enterprise level, data integration infrastructure and governance tools determine which organisations can leverage their data assets. Companies building these analytic capabilities will define what AI applications become trustworthy enough for mission-critical use.
European Strategic Advantage
Where Sovereignty Actually Matters
As the Data Dividend materialises, Europe’s push for technological independence must focus on areas where we have genuine capability or identity-driven imperatives:
In instrumentation, Europe already leads. ASML’s monopoly in EUV lithography reflects a culture of patient, high-precision engineering that naturally extends to data instrumentation. Value is created not through disruption alone, but through integration, systems thinking, and execution alongside industrial incumbents.
Space access is an identity-driven rebuild where European history demands action. Independent access to orbit underpins economic and strategic autonomy. Europe is approaching an inflection point as multiple launch companies near first flight. Backing champions such as our portfolio companies Isar Aerospace and The Exploration Company mobilises Europe’s engineering base toward a resilient, sovereign space infrastructure.
In defence, urgency favors commercial-first innovation. Coordinated procurement remains unlikely. Instead, companies with real advantages will win orders across borders through product-market fit, as Quantum Systems demonstrates. Defence acts as leverage and a short-to-midterm commercial booster for deep tech companies, not their primary market.
In compute, 2026 marks the year Europe shifts from rhetoric to execution. Capital remains the bottleneck. Building competitive silicon requires the entire industrial stack. Without long-term capital backed by coordinated commitment, European compute sovereignty risks remaining aspirational. The repricing of time horizons, recognising 8–10 years as rational, creates a narrow but real investment window.
Regulatory as Competitive Advantage
Europe’s lack of entrenched data monopolies enables radical approaches others cannot pursue: we have no data monopolies to protect. This allows us to pursue radical approaches that the US cannot afford. GDPR forced privacy-preserving architectures years ahead of global requirements. MDR regulations established evaluation standards that Tech Bio now needs worldwide. These frameworks built the data governance infrastructure that makes European systems inherently trustworthy.
Europe can go further. Open, decentralised, privacy-preserving architectures are feasible precisely because no incumbents must be protected. In other words, our distributed structure creates natural testbeds for federated systems. This necessitates data architectures that unlock value across organisational boundaries—the critical enabler for AI-driven enterprise transformation. Companies that master cross-organisational data integration will drive the next wave of enterprise productivity gains, turning Europe’s fragmentation from obstacle into advantage.
For European founders, the goal is not Silicon Valley growth rates or Shenzhen unit economics. It is identifying and dominating bottleneck technologies where precision, reliability, and trust matter more than solely scale—the combination of all is foundational for these. This is Europe’s Lean Production advantage.
The Year of Delivery
As we enter 2026, we are seeing first signs of harvest: robotics achieving scale, enterprise AI gaining real adoption, European space ventures demonstrating genuine capabilities. The gap between AI promise and measurable defines the opportunity.
For Vsquared, the Data Dividend is a core dimension of our deep tech thesis. We invest in the enabling infrastructure: evaluation frameworks and integration tools that make AI trustworthy and productive, molecular instrumentation platforms that generate higher-quality biological data, and industrial robotics that address demographic necessity.
As technical risk declines, founder quality becomes the binding constraint. Europe now produces teams that combine scientific depth with commercial acumen.
The Data Dividend rewards those who control the transformation layer—who turn noise into signal and potential into reality. Like Lean Production before it, this shift is about structure rather than scale. The ultimate validation is workflow redesign: customers changing how they operate because the technology creates undeniable value.
Some of our companies are already there. Many more will follow. For those building the infrastructure rather than chasing headlines, 2026 will be remembered as the year when long-term capital and industrial depth proved their worth.
The question is not whether the Data Dividend will be paid, but who will collect it.





Regarding the topic of the article, your data dividend vision is truly brilliant; I wonder what dynamic new intelligence emergs from such widespred deployment.