Next steps for Australia’s defence innovation: Lessons from DARPA

Australia’s defence innovation community is undergoing a once-in-a-generation reform of its research and development activities, offering a unique opportunity to maximise domestic capabilities. Recent changes¹ to the Defence Science Technology Group (DSTG) — the government’s lead agency for applying home-grown science and technology to national security applications — emphasise external engagement and public-private partnerships under what it ambitiously calls the “DARPA model”.²

The United States’ Defense Advanced Research Projects Agency (DARPA) is arguably the world’s most successful research and technology agency. It is responsible for the development of the internet, the chipsets enabling mobile communications, stealth, night vision, and autonomous vehicle technology. There is much to learn from this organisation in attempting to gain benefits for defence from the advanced R&D activities taking place across Australia. The Australian Department of Defence’s emphasis on external engagement is a welcome starting point. From here it must also address what kinds of external activities DSTG should support and what the optimal mechanisms are by which to support them. A “DARPA model” for Australia can only be achieved through significant reform within Defence to address structural issues that could otherwise produce a gap between expectations and capabilities.

International models for defence research

The 2030 Strategic Plan for the Australian Innovation, Science and Research System explores how Defence — and government more broadly — should interact with the technology sector to drive and better support “innovation”.³ This interaction is critical for meeting Australia’s future security challenges. The economic, national security, and military benefits of scientific R&D are increasingly well understood by governments globally, with China and other Indo-Pacific nations launching their own DARPA-like agencies.⁴ Australia’s 2016 Defence Industry Policy Statement⁵ makes it clear more can be done to foster technology development through better external engagement between Defence and the research sector. But questions remain over how best to achieve this interaction and whether existing civilian programs that fund basic research, like the Australian Research Council, are sufficient for the task.

In contrast to Australia’s civilian-oriented system, the US Department of Defense and military play a central role in supporting scientific discovery and the development of new technologies.

Analysing the US experience in public sector research support may be instructive. In contrast to Australia’s civilian-oriented system, the US Department of Defense and military plays a central role in supporting scientific discovery and the development of new technologies. Although emphasis is placed on the contributions of industry to the domestic technology agenda, the US government frequently acts to underwrite risk, provide early stage support for new ventures, and directly fund private-sector R&D activities. Government funding is provided through civilian basic science agencies,⁶ but also via an alphabet-soup of agencies within Defense and the intelligence community. The government separates “in-house” military research labs, such as the Army Research Laboratory, from agencies that focus on funding external projects. These organisations are known as “Mission Agencies” because they support high-tech R&D outside defence that is focused on military end-user problems, like developing new ways to keep soldiers warm or pack the performance of the most stable atomic clock onto a microchip.

Researchers submitting proposals to these agencies come from academia, national labs, industry, or startups. Funding available from US defence agencies generally outstrips what civilian counterparts, like the National Science Foundation, can offer. The high value of research grants and contracts makes defence a lucrative sector for scientists to submit competitive proposals, and encourages many of the most successful members of the US technical community to apply themselves to defence-related research.

Simply put, basic science is considered a critical area for US defence funding agencies.

In addition to focused R&D, US defence research agencies place a heavy emphasis on basic science and exploratory technology development. For example, the US Naval Research Laboratory has specific divisions on Marine Geosciences,⁷ Oceanography,⁸ and Remote Sensing,⁹ in addition to its much broader divisions on Chemistry¹⁰ and Plasma Physics.¹¹ Major agencies funding external basic research exist inside each military service and include the Office of Naval Research, Army Research Office, and Air Force Office of Scientific Research. Simply put, basic science is considered a critical area for US defence funding agencies. Moreover, there is an entire independent US defence agency — DARPA — funded at approximately US$3 billion annually¹² which has its mission completely detached from military needs. Widely considered the most successful research funding body in the US system, DARPA personnel conduct no research of their own, but focus solely on engagement with, and leadership of, the technical research community through their role as funders.

DARPA’s focus on the future

Despite being housed in the Defense Department, DARPA’s mission is not to develop weapons or enable military superiority. Founded in 1958 in the aftermath of the Soviet Union’s Sputnik launch, DARPA (then ARPA) was tasked exclusively with “preventing technological surprise”.¹³ Sputnik caught the United States off-guard and the government vowed to prevent a repeat by ensuring it invested in the most forward-looking scientific and technological research. Accordingly, DARPA’s annual budget is spread across multiple projects ranging from fundamental mathematics and biotechnology to aerospace systems development.

DARPA is structured around offices that pull together thematically-related research efforts and occasionally change to reflect emerging areas of interest.¹⁴ Within these offices, program managers bring new research ideas and are responsible for managing R&D initiatives and significant budgets. These individuals “pitch” a vision for new research directions to DARPA’s management and, if successful, are empowered to develop funding initiatives that engage the entire innovation ecosystem, from startups to universities.

DARPA’s focus on preventing technological surprise rather than short-term development of deployable technology has enabled the agency to generate totally new classes of transformational technologies.

DARPA pursues this broad portfolio by letting outside experts lead it from the inside. Program managers are not career public servants; they are short-term employees seconded from their roles as academics, industrial engineers, scientists, or other related technical appointments. They typically stay for three to five years and then return to their former positions or move to new roles. This fluid staffing structure accomplishes several intersecting goals: it ensures that researchers bring their best and newest ideas directly to Defense; that groupthink and stagnation are prevented via frequent staff turnover; and that agency management maintains the closest possible contact with R&D.

Regardless of DARPA’s director, the agency exhibits limited top-down strategic management. New staff join on a regular basis, bringing their own views of emerging areas of interest and reshaping the agency’s project portfolio. By virtue of DARPA’s generous funding, new scientific leaders in the broader innovation ecosystem are nurtured. While DARPA’s programs are mostly applied in nature, they rarely address tactical military imperatives directly. Instead, they focus on activities at the “extreme frontiers of physics, chemistry, [and] biology”, such as multi-decade initiatives in basic neuroscience and foundational mathematics.

The results of this approach have been extraordinary.¹⁵ DARPA’s focus on preventing technological surprise rather than short-term development of deployable technology has enabled the agency to generate totally new classes of transformational technologies. For instance, the “DARPA Grand Challenge” and its successors began supporting long-term projects in autonomous vehicles at a time when self-driving cars were largely confined to science fiction films. This began in the early 2000s and continued through challenging military engagements in Afghanistan and Iraq — spanning multiple presidents and producing follow-on challenges that have transformed the field of autonomous vehicles. Such focus is not possible in an agency that needs to rapidly respond to short-term warfighting requirements.¹⁶

Not everything is a success, which is an important part of DARPA’s overall strategy. Projects are driven by technical milestones and funding allocations are reconsidered by senior management approximately once a year. There is flexibility in meeting objectives, but projects that routinely fall short of mutually agreed — and frequently revised — milestones are ruthlessly cut. The oft-stated “DARPA Hard” ethos is that a large fraction of programs need to fail to ensure that the problems being addressed are sufficiently ambitious.¹⁷

The new defence innovation system in Australia

The leading research and development agency in the Australian defence community is DSTG. It has served as a one-stop-shop tasked with everything from advising Defence on the impacts of scientific progress to performing in-house technology and systems integration. The provision of external research support has played a relatively small role in DSTG’s activities from a budgetary perspective.¹⁸ This differs from DARPA’s singular focus on supporting external research and the broader separation of service laboratories from external funding agencies in the US system.

Australia has a history of major defence technology successes — such as over-the-horizon radar and Cubesat surveillance satellites — which Defence has supported by funding university research on a largely ad hoc basis.¹⁹ Nonetheless, the depth of integration between Defence, universities, and industry, compared to the United States, leaves considerable room for improvement. This appears to arise from an institutional focus on ensuring technology investments respond to near-term demand. Indeed, a number of senior Australian military leaders have described the need for technological innovation in terms of meeting contemporary warfighter requirements, rather than focusing on building strategic advantage over the long term.²⁰ The 2016 Defence White Paper, recognising this shortfall, explicitly identified a need for greater interaction between Defence and the research sector in order to embed an innovative culture within the Australian defence community.

Following a series of reviews, Defence has recently implemented a number of significant changes to its own role in the advanced science and technology space. The new defence innovation system has introduced several new entities.²¹ R&D support is now broadly split between the Defence Innovation Hub, focused on late-stage technology development, and the Next Generation Technology Fund (NGTF), focused on defence-related research. The Defence Innovation Hub, managed by the Defence Industry Policy Division, has responsibilities for facilitating research efforts “from concept exploration and technology demonstration, through to prototyping and integrated capability demonstration and evaluation”.²² It has a nominal funding allocation of AU$640 million over ten years. The NGTF is overseen by DSTG and has been allocated AU$730 million over the same period. Sitting alongside these programs is the Centre for Defence Industry Capability (CDIC), which “provides advice to the Australian defence industry, supports industry growth, and facilitates innovation”.²³

Across the defence innovation system there is little space for basic research funded by the Department of Defence.

Three things emerge on reviewing these new initiatives. First, both the Defence Innovation Hub and CDIC are primarily focused on supporting industry. CDIC, for instance, is run by AusIndustry and does not list scientific, technological, or strategic outcomes within its mission or vision statements. These two industry-subsidising organisations are, in combination, funded above the level of Defence research activity through the Next Generation Technology Fund.²⁴

Second, the emphasis on technology investment that does exist has a fairly traditional focus on military needs. The “innovation priorities”²⁵ that set overarching guidance for both the Defence Innovation Hub and NGTF identify six streams according to the Force Structure Review: intelligence, surveillance, reconnaissance, electronic warfare, space and cyber; key enablers; land combat and amphibious warfare; strike and air combat; maritime and anti-submarine warfare; air and sea lift.²⁶ Even the “strategic priorities” of the new NGTF remain biased towards traditional defence technologies, with emphasis on: directed energy weapons; intelligence, surveillance, reconnaissance; cyber and space.²⁷

This portfolio is not fully consistent with best practices in DARPA and exposes an opportunity to better capitalise on a large fraction of Australia’s research ecosystem. Across the defence innovation system there is little space for basic research funded by the Department of Defence. Only one “addendum” category of the Defence Innovation Hub lists “unexpected or unforeseen technologies or ideas”.²⁸ As such, there is no clear funding category for fundamental non-applied research, and only limited provision for “forward looking” research into general topics of interest to Defence, such as quantum technologies and material science.²⁹

Third, the new structure for managing external research support adds this activity to existing administrative structures rather than creating a dedicated agency focused exclusively on external research. DSTG, while overseeing the NGTF, continues to maintain “core” roles in acquisition, sustainment, and operations.³⁰ Only one of ten articulated roles for DSTG relates to external research partnerships, and publicly available materials emphasise its provision of support to Australian military operations. This broad mission for DSTG not only draws attention and resources away from developing cutting-edge research, but also introduces potential conflicts of interest in how and to whom it awards funding. The organisation has previously been criticised as “DSTG-centric” in its approach to enabling advanced technology development for Defence, with most of its budget supporting staff expenditures. This may also apply to the award of competitive research funds.

Australia’s need for home-grown tech innovation

Australia’s R&D sector is a national asset that should be used for both economic and strategic advantage. There has been much debate over the need for an indigenous defence industry capable of producing large-scale military weapons systems, but relatively little discussion over the value of securing domestic support for research, including from the entities in Australia’s new defence innovation system.

Drawing on the DARPA experience, Australia’s defence innovation system has an opportunity to better support domestic industry through higher external research investment. DARPA’s experience challenges the assumption that dedicated warfighting or industry-supporting entities must be embedded in a research funding system. The capabilities developed by DARPA have spawned new industries and technologies of their own.³¹ These have not only indirectly led to advanced military technologies; but have also enhanced domestic R&D activity more broadly, providing new opportunities for capacity-building and value capture. The context for this type of R&D dispersion in Australia is ripe. According to the chief scientist, 26 per cent of all current economic activity has arisen from basic scientific advances over the last 20-30 years.³² In other words, the whole economy would benefit from greater investment by Defence into advanced research.

Drawing on the DARPA experience, Australia’s defence innovation system has an opportunity to better support domestic industry through higher external research investment.

The strength of Australia’s quantum computing sector shows how targeted defence investment can facilitate major domestic technological and industrial successes. Supported by US military and intelligence agencies — which have invested tens of millions in Australia since 2000 — Australia’s quantum sector is now recognised for its contributions to developing useful quantum technologies.³³ Canberra, in this case, has benefitted from the benevolence of the US government as an allied “Five Eyes” nation. But there is no reason that similar investments could not be domestically funded. Indeed, political changes in Washington suggest that relying on large-scale US investment for future research initiatives may be unwise.

Higher investment by Defence in advanced technology is needed. The NGTF, at AU$73 million per year, constitutes just 0.2 per cent of annual defence expenditure; compared with the approximately two per cent of US defence expenditure on science and technology, or the 0.5 per cent allocated solely to DARPA.³⁴ Scaling DARPA’s budget based on Australia’s population would yield AU$300 million per year, four times the NGTF. This would provide for a major shift in the way the Australian Department of Defence supports advanced R&D. For instance, funding individual projects at AU$2 million annually for five years — roughly the same as a typical DARPA project — would enable about 150 major new technological initiatives to begin, with funding approximately ten times higher than what is available from existing scientific funding bodies.

Next steps for Australian defence innovation

Australia must continue to diversify the sources of fundamental research funding to more routinely include Defence as a major driver of innovation. DARPA’s success highlights the role that incentivised competition amongst the top performers in a research ecosystem can play in developing new technologies of national interest for both defence and civilian applications. Australia’s defence innovation system needs a standalone DARPA-like agency defined by three primary characteristics:

1. Outward focus: The new agency should be dedicated to supporting research by partners in industry, government, and academia, rather than supporting internal R&D performed by defence scientists. In line with DARPA, approximately 90 per cent of its budget should be directed to external programs.
2. Expert rotational technical staffing: The agency should be staffed by technical experts who can drive new research programs, mitigate stagnation, and maintain an ongoing connection between agency management and active researchers. Ideally, rotations with academic and industrial research organisations should be supported, but career public servants may also rotate through the agency and all should incur term limits.
3. Broad strategic portfolio separated from operational needs: Like DARPA, the agency needs a broad portfolio ranging from basic science and mathematics to tactical technology development, and must be divorced from contemporary operational needs within Defence. This is essential to maintain the long-term vision of such an agency. Segregation of internal R&D missions can mitigate conflicts between near-term military imperatives and long-term strategic initiatives.

A clearly articulated and narrowly scoped vision statement focused on supporting the most forward-thinking R&D activities is essential to ensuring such an agency can flourish. An Australian DARPA needs to focus exclusively on solving problems deemed by technical experts to be important in maintaining a scientific and technological advantage. The potential military relevance of research projects can be considered, but should not necessarily be a primary driver of investment. To spawn genuine innovations in emerging technologies, the portfolio must contain long-term and high-risk activities that may never yield deployable systems.

Establishing an agency with the statutory authority to engage in high-risk, high-reward research is an essential missing piece of Australia’s new defence innovation strategy.

DARPA’s leadership makes clear that embracing the possibility and even likelihood of failure is essential to the mission.³⁵ Aggressive government scrutiny of public funding for research breeds a fear of failure that can constrain research agencies and, in turn, the entire research enterprise. Risk management, rather than risk aversion, is required. Establishing an agency with the statutory authority to engage in high-risk, high-reward research is an essential missing piece of Australia’s new defence innovation strategy.

Boosting the role of Defence in Australia’s innovation ecosystem requires broad structural, cultural, and tactical changes. Australia’s R&D sectors are strategic assets that can be better leveraged for national benefit with the right form of engagement from the defence sector. Just as Canberra accepts the need for long-term force modernisation, the Department of Defence should embrace a broader vision for the development of advanced technologies that will position Australia at the forefront of new technological capabilities. The first steps taken in establishing the NGTF put Defence on a pathway to capitalising on domestic research strength. It is time to complete the journey by implementing the DARPA model in full.