From Bits to Atoms. Why Venture Still Avoids the Industries That Run on Physical Infrastructure – 0100 Weekly Brief
Hello there!
Venture investors are more and more interested in Deep Tech, but recent data shows a growing disconnect between where capital flows and where real-world deployment is needed.
So, what better time to look at why the industries that run on physical infrastructure are still being overlooked?
DeepTech now represents 32% of all VC funding in Europe, and the total investment reached $20.3bn in 2025, only 4% below the 2021 peak, while traditional tech remains 54% below its peak. On the surface, capital is moving toward science-driven innovation.
However, allocation is uneven.
A large share of recent growth is concentrated in AI-related segments, particularly software-driven or compute-heavy layers. In contrast, capital flowing into sectors tied to physical infrastructure, energy systems, industrial automation, and robotics deployment remains smaller and more volatile.
This creates a split inside Deep Tech: capital flows into “bits” (AI, compute, software-enabled science), while “atoms” (energy, manufacturing, construction, robotics in the field) require longer timelines and still struggle to attract consistent funding.
AEC-Tech provides a concrete example. Despite being one of the largest global industries, it accounts for only 0.3%–0.6% of total VC funding. Even with $40bn in cumulative VC investment, its share remains marginal relative to the software sector. Capital is entering Deep Tech, but it is not flowing in proportion to the industries where physical systems dominate.
Construction robotics also provides a useful reference point for how early this change still is. According to a recent report by Zacua Ventures, on-site robotics represents less than 0.03% of global construction spend, despite being one of the most discussed areas. Even with steady growth at mid-teens annual rates, the market remains a small base relative to the size of the underlying industry.
The Industries That Matter Most Are Still Physical
Deep Tech is positioned as the solution to global challenges, including climate change, resource scarcity, and demographic decline. These are not software problems. They require deployment in energy systems, industrial production, and built environments.
According to Dealroom, electrification is increasing demand for critical materials like lithium and rare earths, while demographic shifts are driving the need for automation in sectors like manufacturing and construction. These are capital-intensive systems tied to physical infrastructure.
In construction, this translates into a set of clearly defined workflows where automation is emerging. These include layout and measurement, groundworks, structural rebar, and inspection, which already account for most of today’s robotics revenue.
Tasks that are repetitive, physically demanding, and directly tied to project timelines, such as trenching, layout marking, or rebar tying, are the first to be automated. This shows that adoption is not related to technological changes but is driven by operational challenges.
Europe Has The Assets, But Not The Capital Structure
According to Dealroom and a few other market reports, Europe produces 1.5 million STEM graduates per year and employs 2.15 million researchers, with 21.8% of global patent applications.
It also hosts 30% of the world’s top Deep Tech universities. The supply side is not the constraint.
Besides this, Europe has a massive advantage when it comes to expertise. Almost 80% of VC-backed Deep Tech founders who have built a company after 2020 are highly educated and have a technical background.
The challenges come when we talk about capital allocation and scaling. European Deep Tech startups face an annual funding gap of $4bn to $24bn, especially at later stages. As a result, 70% of late-stage funding comes from non-European investors, often shifting companies’ center of gravity abroad .
This matters more for “atoms” than “bits.”
Physical infrastructure companies require sustained capital over longer time horizons. When that capital is unavailable locally, companies either under-scale or relocate.
This gap is amplified in AEC-Tech.
While total funding reached $550m in Q4 2025, it was still described as one of the weaker quarters and reflects volatility rather than consistent capital scaling. The issue is not initial funding but the continuity of capital over long deployment cycles.
The Real Bottleneck is Not Innovation, It is Industrialization
Deep Tech companies require upfront CapEx and longer development cycles before they generate revenue, which entails different risk profiles that investors need to understand to build their portfolios.
In AEC-Tech, this is visible in how solutions are adopted. Innovation spans on-site robotics, modular construction, building materials, and supply chain platforms. But adoption depends on integration into fragmented ecosystems, contractors, developers, regulators, and suppliers.
The funding data reflects this friction. Even as Series A/B rounds stabilize and late-stage rebounds, there are no net-new unicorns in Q4 2025. Capital is available, but scaling outcomes remains limited.
The venture model is optimized for speed and iteration. Physical systems require coordination, infrastructure, and time. This mismatch explains why capital concentrates upstream rather than downstream.
Sovereignty is Driving Investment, But Mostly in Strategic Sectors
Geopolitical pressure is accelerating investment in certain Deep Tech areas. Defence, security, and resilience accounted for 43% of Deep Tech funding in 2025, up from 20% in 2022. According to Dealroom, public capital is scaling across strategic sectors like:
€43bn EU Chips Act
€109bn France AI and data infrastructure
€35bn Germany space investments
€20bn+ annual defense-linked Deep Tech spend
These sectors are capital-intensive and physical, but they benefit from coordinated public demand.
AEC-Tech does not. Despite being essential infrastructure, it lacks equivalent policy-driven capital flows. The report shows EU + Israel lead globally in AEC-Tech, but still with only ~0.3% share of global VC
From Bits to Atoms: Where The Next Returns May Come From
As software becomes easier to build and replicate, defensibility shifts toward systems that combine software + hardware + infrastructure. The report notes that falling software costs are pushing investors toward “the physical and fundamental.”
Historically, major value-creation cycles, from telecom to manufacturing automation, have originated from shifts in physical infrastructure. For example, the cost of data transmission dropped 1000x over 30 years, and manufacturing time per vehicle declined from 40h+ to 10–15h. AEC-Tech shows early signs of this trend. Funding remains above pre-2021 levels (+49% on average per quarter) despite volatility.
The next cycle, energy systems, robotics, industrial automation, and construction, follows the same pattern. The question is - are venture investors ready to adopt the right model to capture it?
📝 Expert Interview | Inside the $25 Trillion Future of the Built Environment
To understand how central these industries are, it helps to look at their scale. The built environment, covering construction, real estate, infrastructure, and energy systems, is estimated at over $25 trillion globally, making it one of the largest sectors in the world. Yet despite this scale, it has seen limited technological transformation over the past decades, largely because of the way risk and incentives are structured across the value chain.
As Juan Nieto, General Partner at Zacua Ventures, explains, adoption is not just a technical challenge but an incentive problem. In construction, risk is often pushed down to contractors and subcontractors, meaning that adopting new technologies can directly conflict with short-term project economics.
Let’s Continue the Conversation at 0100 Europe
At 0100 Europe, we’ll take this discussion further in a dedicated panel on “Investing in the Next Frontier: What Makes Deep Tech Worth the Risk?”, focusing on how venture capital is approaching Deep Tech today, and what actually makes these investments work from early-stage innovation through to market impact.
The conversation will look at where capital is flowing, which segments are gaining traction, and how investors are navigating longer timelines, higher capital intensity, and the realities of scaling beyond the lab.
Joining the panel are investors actively shaping this space across different markets and strategies, including Ben Scheidt (Redstone), Patricia Rinke (AfricInvest), Kirsten Connell (Octopus Ventures), Amelia Armour (Amadeus Capital Partners), and Rubina Singh (Foresight Ventures). Together, they bring perspectives across AI, sustainability, early-stage Deep Tech, and emerging ecosystems