In one television conversation with an experienced engineer, I once heard a strikingly accurate formulation: “You cannot sell a technology if you have nowhere to show the process.” He explained that new solutions used to be demonstrated in a factory, where a client could see how materials behaved, how the process unfolded, and where real disturbances appeared. Today, such spaces no longer exist in many places, and as a result, there is nowhere to turn an idea into a convincing technological offer. This insight is not nostalgia for the past—it reflects the logic of contemporary innovation: without a physical demonstration in the middle segment, a technology remains nothing more than an abstraction.
More recently, however, an entrepreneur offered almost the opposite perspective: “Scientists should come to my company and learn — I don’t have to pay for that.” It is a different position, yet it exposes the same gap. In this view, the entrepreneur sees himself as a passive observer rather than a partner in the making of innovation. It also reveals a deeper structural problem: there is no mechanism to cover the costs of learning, testing, or jointly defining a technology. “Learning” is treated as a private burden of the scientist rather than an integral part of the engineering process.
This produces a systemic contradiction: the state expects scientists to deliver market-ready results, the market expects scientists to self-organise, yet few are willing to finance the very segment where this transition actually happens — the middle ground between science and industry. The result is a structured void that blocks the possibility of innovation regardless of a country’s size or economic strength. Technologies do not emerge from individual will or good intentions: they arise from a professionally organised, financed, and engineering-equipped collaborative space where ideas can become processes.
What Is Innovation?
Joseph Schumpeter (1934) drew a very clear distinction between two phases that are often conflated: invention, the creation of an idea, and innovation, its implementation. An invention by itself does not change the world; it becomes an innovation only when a technology is tested at real scale, demonstrated to a client, integrated into a concrete production process, and ultimately generates new economic structures. This is why countries where innovation proceeds smoothly invest not only in scientific discovery but also in the implementation stage — the complex and costly arena where an idea becomes a working process.
In contrast, in countries where innovations tend to stall, there is often an expectation that the market will “solve everything” on its own. Yet market logic does not cover the risk between the laboratory and the factory: it is an expensive, technologically demanding, and unpredictable phase whose maintenance cannot be delegated to individual firms. Schumpeter described this divide between creation and implementation as the implementation gap. Today it corresponds to the TRL4–7 levels, and this is precisely where most good ideas “die”: not because they are flawed, but because they lack the space in which to become real. The middle segment is missing.
Where Has the Middle Segment Gone?
The middle segment is the heart of the technological process — the space between science and the market that is not immediately visible, yet this is precisely where everything decisive happens. In countries where innovation is not a rarity, a carefully maintained and cultivated bridge always exists between the laboratory and commercial deployment: pilot production, scaling centres, demonstrators, engineering teams, and safety or quality validation procedures. This bridge is not a decorative accessory — it is the mechanism that allows an idea to become a usable, functioning, and comprehensible process ((Fraunhofer-Gesellschaft, 2024; OECD, 2025).
When this segment is missing, science remains abstract and unsuitable for the realities of firms; meanwhile, industry remains cautious because it has no evidence that ideas will work on its equipment. And the market? The market has nothing it can evaluate, measure, or purchase. An idea cannot be sold like a book or a piece of software — it can only be sold once it is real, when it can be switched on, touched, and tested.
The engineer’s remark in a television interview — “I have nowhere to demonstrate it” — is a global symptom. International literature shows quite consistently that a structural “valley of death” exists between early research and market adoption in many countries — a zone where middle-segment (TRL 4–7) projects are underfunded, and thus a portion of potentially viable technologies never reach industry (Ellwood et al., 2020; Gbadegeshin et al., 2022; Kampers et al., 2021). Similar conclusions appear from Canada (Innovation, Science and Economic Development Canada, 2021) to India, where policy analyses identify funding gaps as one of the central systemic barriers to innovation (Council on Energy, Environment and Water & Shakti Sustainable Energy Foundation, 2024; Crocker et al., 2023).
In Latvia, universities and research institutes do possess individual middle-segment elements — prototyping workshops, innovation centres, or pilot lines in narrow niches. However, taken as a whole, these form isolated “islands, micro-islands, or archipelagos,” not a systemic, nationwide middle segment capable of linking science and industry at scale. Ideas requiring specific scaling often hit the same barrier: there is no place to test them under pilot-production conditions (OECD, 2021; European Commission, 2018).
Technology as a Form of Social Relations of Production
Within the tradition of critical philosophy of technology, convincingly developed by the American philosopher Andrew Feenberg, technology is neither an instrument nor a neutral artifact. It is a socially constructed, action-enabled network in which specific power relations, institutional regimes, and collective capacities become embodied. Technology emerges only when a society possesses a complete socio-technical configuration: engineering competence, financial mechanisms for piloting and testing solutions, infrastructure that allows those solutions to be examined in practice, and stable forms of collaboration between science and industry. Feenberg emphasises that without such a configuration, technology has nowhere to materialise: it remains an abstraction with no opportunity to become embodied in a real process (Feenberg, 2002).
This is not an individual or local deficiency but a structural deformation characteristic of systems in which the institutional layer of the middle segment is weak or absent.
Can the Middle Segment Be Replaced by Digital Twins?
In contemporary innovation discourse, one often encounters the hope that digital twins, simulations, or video demonstrations might replace the middle segment of the physical world. Yet from the perspective of critical philosophy of technology, this is a misleading assumption. Digital twins are valuable for optimisation, but they cannot embody what Feenberg calls the socio-technical reality of technology: material irregularities, chemical risks, unpredictable boundary phenomena, and the true costs that become visible only during scaling. Digital twins are a powerful tool for optimisation and scenario analysis, but both engineering and policy literature consistently describe them as complements — not substitutes — for physical experiments and pilot production. As model complexity increases, the range of questions that can be resolved only in the physical world expands rapidly. At the systems level, public institutions also point to the challenges digital twins face in terms of scalability, data requirements, and costs, which prevent them from replacing physical infrastructure (National Academies of Sciences, Engineering, and Medicine, 2024; Johra et al., 2021; Government Accountability Office, 2023).
For an investor or an industry partner, a description, narrative, or video is not enough. They want to see a process that works on their equipment, under their operating conditions, and under their control. A simulation cannot provide that proof because it does not engage the real engineering structure — only a modelled derivation of it. Digital twins, therefore, are not an absolute substitute for the middle segment but a supporting instrument.
Innovation Is a Chain
When we speak about innovation, we often think of an idea — a spark, a moment of inspiration, or a creative insight that seems, by itself, capable of producing a new technology. But in reality, innovation is neither a moment nor an image; it is a chain in which every segment is necessary for an idea to enter the material world at all.
There are no superfluous elements in this chain. Research without pilot production remains an abstraction. Industry without science has no direction. Investment without engineering cannot turn into a process. And even the most precise digital models cannot replace what happens only in the physical world — where materials respond, where tension arises, where technology reveals its true nature.
This is why the countries in which “unicorns” consistently emerge are not those with the most ideas. (That is a romantic illusion.) Innovation happens where the entire chain has been built to completion: where there is space for demonstration, space for error, space for material experimentation, space for engineering teams (who know not only how to imagine, but how to build), and space for institutions and entrepreneurs willing to assume risk and responsibility, instead of waiting for someone else’s courage.
References
Council on Energy, Environment and Water, Shakti Sustainable Energy Foundation. (2024). The India Climate Finance Report: Landscape and flows. https://www.ceew.in Accessed: 25 November 2025
Crocker, S., Konduru, N., Chuttani, G., Paul, O., Tambos, E. (2023). Improving India’s innovation system through reformation of the technology-based start-up sector (GLA1011H Student White Paper 2023-01). Innovation Policy Lab, University of Toronto.
Ellwood, P., Williams, C., Egan, J. 2022.Crossing the valley of death: Five underlying innovation processes. Technovation, Volume 109, 102162. https://doi.org/10.1016/j.technovation.2020.102162.
European Commission. 2018. The Latvian research funding system – Specific support to Latvia – Horizon 2020 Policy Support Facility, Publications Office. https://data.europa.eu/doi/10.2777/3392. Accessed: 25 November 2025.
Feenberg, A. (2002). Transforming technology: A critical theory revisited. Oxford University Press.
Fraunhofer-Gesellschaft. (2024). Finances. Fraunhofer-Gesellschaft. Accessed: 25 November 2025
Gbadegeshin, S.A., Al Natsheh, A., Ghafel, K., Mohammed, O., Koskela, A., Rimpiläinen, A., Tikkanen, J., Kuoppala, A. 2022. Overcoming the Valley of Death: A New Model for High Technology Startups. Sustainable Futures, Volume 4, 100077. https://doi.org/10.1016/j.sftr.2022.100077.
Government Accountability Office. (2023). Science & tech spotlight: Digital twins. U.S. Government Accountability Office. https://www.gao.gov/products/gao-23-106453. Accessed: 25 November 2025
Innovation, Science and Economic Development Canada. (2021). Evaluation of the Strategic Innovation Fund (SIF). Government of Canada. https://www.ic.gc.ca/eic/site/ae-ve.nsf/eng/h_03942.html Accessed: 25 November 2025
Johra, H., Aleksandrova Petrova, E., Rohde, L., Pomianowski, M.Z. 2021. Digital Twins of Building Physics Experimental Laboratory Setups for Effective E-learning. Journal of Physics: Conference Series. 2069, 012190. DOI: 10.1088/1742-6596/2069/1/012190
Kampers, L.F.C., Asin-Garcia, E., Schaap, P.J., Wagemakers, A., dos Santos, V.A.P.M.
Navigating the Valley of Death: Perceptions of Industry and Academia on Production Platforms and Opportunities in Biotechnology. EFB Bioeconomy Journal, Volume 2, 100033. https://doi.org/10.1016/j.bioeco.2022.100033.
National Academies of Sciences, Engineering, and Medicine. (2024). Foundational research gaps and future directions for digital twins. National Academies Press. https://www.ncbi.nlm.nih.gov/books/NBK605519/ Accessed: 25 November 2025
OECD (2021), “Innovation Diffusion in Latvia: A Regional Approach”, OECD Regional Development Papers, No. 83, OECD Publishing, Paris, https://doi.org/10.1787/7d6d0ffc-en. Accessed: 25 November 2025
OECD. (2025). Regional impact of public R&D organisations. OECD Publishing. https://www.oecd.org/en/publications/regional-impact-of-public-r-d-organisations_587d7e6c-en.html Accessed: 25 November 2025
Schumpeter, J. A. 1934. The theory of economic development. Harvard University Press.
Technology Startups. Sustainable Futures, Volume 4, 100077. https://doi.org/10.1016/j.sftr.2022.100077.
empiricux.com 