Making the Invisible Visible: Software as Strategic Infrastructure in the Australian Economy

John H Howard, 27 June 2025

Software is no longer just a tool or a tech sector — it is the invisible infrastructure powering Australia's physical economy. From energy and health to defence and logistics, software and AI now underpin national productivity, coordination, and resilience.

This Insight calls for a strategic reframing of software as critical infrastructure, essential to building sovereign capability and unlocking long-term economic performance.

---

Software has long been misunderstood in the framing of innovation policy. Although it is more or less visible in mobile apps and consumer platforms, software now represents the invisible architecture of Australia's industrial and economic systems.

Software enables the integration of machines, data, work and workflows, as well as decisions, across domains as diverse as mining, healthcare, transportation, manufacturing, energy, and defence. In this new configuration, software is not just a tool; it is the infrastructure on which modern industrial capability rests.

This Insight argues that public policy must shed outdated conceptions of software as a self-contained "tech sector" and instead recognise it as a critical, yet largely hidden, enabler of Australia's physical economy. It explores software's embedded role in industrial competitiveness, its systemic but under-recognised contribution to productivity growth, and the need for deliberate national capability-building to ensure its long-term contribution to economic resilience and performance.

Rethinking the Role of Software in the Economy

While most people encounter software through visible interfaces—such as apps, digital services, and user-facing platforms—these are merely the surface expressions of a far more significant reality. Beneath the interface lies software's more consequential role: that of a foundational infrastructure. This argument was addressed in our Insight of 12 May 2025, Computer/Information Services and Australia’s Path to a Future Made in Australia.

Software enables the integration of machines, data, work processes, and decision logic across systems that were once isolated or manually operated. In doing so, software transforms physical operations into coordinated, responsive, and increasingly intelligent networks. It does not function as a peripheral enhancement but rather as the binding mechanism that enables sectors to operate with higher degrees of precision, adaptability, and scale.

This integrative function is particularly evident across several key domains. In energy systems, software coordinates the generation, storage, and distribution of electricity, facilitating real-time predictive forecasting, grid stability, and load balancing. In mining, software supports orebody modelling, manages autonomous haulage fleets, and underpins remote monitoring and safety systems.

In health and bioinformatics, it drives diagnostic platforms, genomic analytics, and electronic record management, enabling distributed care systems to operate with precision and consistency. In logistics, software governs fleet coordination, optimises multimodal transport, and adapts routing to real-time conditions. In manufacturing, software tools enable simulation, robotic control, and process refinement in high-precision environments.

Beyond these sectors, software is increasingly integral to infrastructure planning, public service delivery, cultural production, and national security. What binds these diverse applications is not simply the use of digital tools but their deep dependence on robust, often invisible software systems that ensure operational reliability, system-level performance, and sustained scalability

Software and Productivity: A Systems-Level Contribution

The contribution of software to productivity is often underestimated, largely because traditional economic metrics are poorly suited to capturing its systemic and intangible impacts. Unlike physical capital, software does not conform to familiar patterns of depreciation. Nor does it resemble labour in its direct and observable output.

Instead, software often works indirectly—embedded in systems, routines, and processes—to reshape how work is done, decisions are made, and resources are allocated. Its influence becomes visible in aggregate improvements, efficiencies, and the creation of new capabilities that would be impossible to achieve through conventional inputs alone.

Software contributes to productivity through several interlocking mechanisms. It automates routine or repetitive tasks, thereby reducing human error, downtime, and resource inefficiency. In doing so, it allows human and financial capital to be redirected towards higher-value functions. It also enhances decision-making by embedding expert knowledge within tools that visualise data, model scenarios, and simulate outcomes, extending the reach of specialist insight across organisations.

Software facilitates the integration of separate and complex systems, allowing disparate functions or fragmented assets to operate as unified, responsive networks. This is particularly critical in domains such as enterprise resource planning (ERP) and customer relationship management (CRM), as well as in logistics, energy, and healthcare, where coordination across time and space underpins performance. Software also enables a degree of adaptability and customisation that traditional technologies cannot match, supporting innovation in service delivery, pricing models, and customer engagement.

In the food industry, software and systems integration are essential to unlocking productivity across the full value chain—from farm to fork. Integrated software platforms coordinate supply logistics, optimise cold chain management, and enable dynamic inventory and pricing systems in retail and food service. In artisanal and multicultural food production, software tools support compliance, traceability, and batch-level customisation.

These capabilities are particularly important as food systems shift from mass commodity output to diverse, experience-driven consumption. Without a robust software infrastructure, these innovations struggle to scale, and the productivity potential embedded in Australia’s broader food economy remains under-realised.

Productivity gains are not, therefore, automatic. They rely on complementary capabilities: robust data governance, organisational change management, skill development, and infrastructure interoperability. In their absence, software remains underutilised, and its productivity potential is unrealised.

The Fusion of Digital and Physical Economies

The term "digital transformation" is often used to describe a shift from physical to digital systems, implying that digital tools replace traditional processes. However, this framing is misleading. In practice, the more accurate description is one of fusion, where software does not displace physical systems but instead enhances, augments, and orchestrates them. This interdependence is becoming the defining characteristic of contemporary infrastructure.

In agriculture, for example, sensor networks embedded in fields transmit real-time environmental data to software platforms that manage irrigation, fertiliser application, and harvest planning–delivering what is often termed precision agriculture. Rather than replacing the act of cultivation, software refines it, improving yield and reducing waste.

In the energy sector, control software and digital twins manage grid balance and coordinate renewable energy sources, enabling dynamic responses to fluctuations in generation and demand. In transport, logistics software connects modes—road, rail, air, and sea—allowing for adaptive scheduling and route optimisation based on traffic, weather, and supply chain conditions.

This fusion alters the logic of investment and governance. Software becomes the operating system of the physical economy—essential to its reliability, performance, and adaptability. Recognising this interdependence invites a new approach to planning and resource allocation–one that treats digital systems as integral to the productivity of land, capital, and human resource assets.

Underrepresentation in Measurement and Policy

Although software plays an increasingly critical role in shaping productivity, coordination, and system resilience, it remains systematically underrepresented in both national accounting frameworks and public investment strategies. This disconnect reflects several persistent structural and institutional limitations that prevent software's true economic significance from being recognised and leveraged.

• The uneven adoption of software across sectors. High-performing industries such as finance and telecommunications often lead in integration, while others—particularly in regional manufacturing, healthcare delivery, and construction—remain under-digitised. These gaps dampen the aggregate productivity benefits that software might otherwise deliver.

• Many organisations lack the system readiness to implement software effectively. Legacy infrastructure, insufficient workforce capabilities, and weak change management practices can hinder the successful adoption of new systems.

• As argued in our Insights posting of 24 June, Modernising Industry Classifications for a Services-Driven Economy, the metrics used in national statistics still prioritise tangible assets and discrete sector classifications. Software, by contrast, is intangible, iterative, and deeply embedded across multiple domains—traits that current frameworks struggle to accommodate.

• Software is no longer just a tool or a tech sector — it is the invisible infrastructure powering Australia's physical economy. From energy and health to defence and logistics, software and AI now underpin national productivity, coordination, and resilience.

Together, these dynamics reinforce a policy bias that sees software as ancillary rather than foundational to economic strategy.

From Fragmented Investment to Systemic Capability: Rethinking the Foundations of Software Infrastructure

Historically, investment in software capability has been fragmented. Funding has focused on start-up support and short-cycle innovation, while skills programs have emphasised basic digital literacy. These efforts have value, but do not address the structural features of software infrastructure: longevity, interoperability, and reliability.

The institutional landscape reflects this fragmentation. Applied research is distributed across universities, government research agencies, and Cooperative Research Centres (CRCs), with few entities responsible for the long-term stewardship of shared software platforms. International exemplars — such as Germany's Fraunhofer Institutes or Singapore's GovTech — offer alternative approaches with mission-driven mandates and sustained public investment.

Treating software assets as infrastructure rather than proprietary tools would change investment logic. Shared libraries, domain standards, and public code repositories can lower barriers to entry, reduce duplication, and ensure continuity in mission-critical domains.

Questions of sovereignty are also reframed under this view. Security becomes a function of capacity: the ability to verify, adapt, and maintain systems independently. Operational assurance matters more than control of ownership. In sectors like defence, mining, and critical health infrastructure, domestic capability in software integration and verification is as important as access to the source code.

Workforce development presents a related challenge. While basic digital skills are widespread, deep technical capabilities in applied software engineering — particularly in areas such as embedded systems, distributed computing, and AI integration — remain limited. Strategic education pathways and domain-specific fellowships may better align training with system-level needs.

Procurement policy also plays an important role. Public sector demand can help anchor markets for secure, standards-compliant software. Infrastructure tenders, in particular, offer opportunities to signal long-term priorities and to support testbeds for integration and system validation.

Amplification by Artificial Intelligence: Recasting the Software Narrative

For decades, software and artificial intelligence (AI) have evolved together, often indistinguishably. The longstanding adage, "this year’s AI is next year’s software," reflects a practical reality: what is initially regarded as advanced AI eventually becomes embedded within routine software systems. Despite their conceptual separation in contemporary discourse, AI is not a distinct technology from software—it is a dynamic extension of it.

Recent advances in AI have reignited public and policy interest in software by surfacing new capabilities—natural language interfaces, predictive modelling, adaptive automation—that appear novel but are built on existing software foundations. In this light, AI has amplified software’s visibility and strategic relevance. It has made software “interesting” again, even though the underlying mechanisms—code, algorithms, data structures—remain part of the same lineage.

This amplification also introduces greater complexity and consequence. Embedding AI into software systems that govern critical infrastructure—such as energy grids, logistics chains, and health diagnostics—intensifies the need for robust safety protocols, transparent logic, auditable performance, and rigorous system governance.

The more embedded and autonomous these systems become, the more important it is that their underlying software is secure, explainable, and engineered for reliability. Trustworthy AI is not a separate achievement but an outcome that depends on the quality and integrity of the software systems in which it operates.

Australia’s digital and industrial strategies must acknowledge that the excitement surrounding AI does not diminish the importance of software—it reinforces it. Investment in AI capability is only meaningful if matched by investment in the software infrastructures that enable, constrain, and scale its application. The future of AI is inextricably linked to the future of software as infrastructure.

Building the Invisible Architecture of a Sovereign Digital-Physical Economy

Australia's economic resilience, industrial capability, and strategic autonomy increasingly depend on assets that extend beyond physical infrastructure alone. Software — and increasingly, software fused with artificial intelligence — constitutes the invisible architecture that binds together machines, data, workflows and workers, and decisions across every major sector of the economy.

To emphasise, software must be reframed as strategic infrastructure: public-purpose, long-cycle, and mission-critical. It enables not just digital functionality but the integration of Australia's physical systems — from grids and ports to hospitals and factories. The convergence of software and AI reinforces this argument.

Building sovereign capacity in these domains demands more than skills programs or digital rhetoric. It requires deliberate public investment in shared software infrastructure, applied research institutions, standards, security protocols, and interoperable platforms.

Capacity building also demands talent pipelines that span software engineering, AI systems, and domain-specific problem-solving. And it requires reform of national accounting and procurement mechanisms to reflect the public value of intangible, embedded digital assets.

Australia's future competitiveness will not rest solely on physical capital, nor even on data itself, but on the software infrastructure — increasingly intelligent, adaptive, and mission-oriented — that allows systems to sense, decide, coordinate, and act.

The opportunity now is to make this invisible infrastructure visible: to recognise it, invest in it, govern it, and ensure it serves the national interest. Without it, ambitions for productivity, sustainability, and sovereignty will remain aspirational. With it, Australia can embed intelligence into the very fabric of its economy and institutions — and meet the challenges of a deeply digital-physical future with purpose and capability.