< Back to Whiteboard

Wicked Game: Multi-Domain Integration and the Problem of Complexity

Hannah Croft
Hannah Croft 5 Aug 2021
Multi-Domain Integration
Wicked problem

Multi-Domain Integration and the Problem of Complexity

A complex system designed from scratch never works.

The fundamental premise of Multi-Domain Integration (MDI) - to ‘sense, understand, and orchestrate effects at the optimal tempo across the operational domains and levels of warfare’ – will be achieved, it is argued, by addressing gaps in information flows and breaking up Command and Control (C2) stovepipes in UK Defence and Partners Across Government.

The desired end state emerges as an agile, dynamic C2 system complimented by a common operating picture that converges information in and from different domains and levels of warfare, where ‘efficient levels of processes, permissions and information exchange capacities’ can orchestrate cross-domain thinking and manoeuvre[1]. Such a complex Command, Control, Communications, Computers, Information/Intelligence, Surveillance, Targeting Acquisition and Reconnaissance (C4ISTAR) system, notes Joint Concept Note 1/20, is ‘fundamental to enable MDI – it is the key requirement’.

How will this system, what the Ministry of Defence (MOD) describes as the ‘linking glue'[3] to MDI, be realised? I’ve recently got into Systems Theory (things happen when you work at Rowden) and it’s probably cliché of me to drop some John Gall on you, but it seems like a pretty neat guiding principle for such a task:

A complex system that works is invariably found to have evolved from a simple system that worked. A complex system designed from scratch never works and cannot be patched up to make it work. You have to start over with a working simple system.

So, extending the MOD’s linking glue metaphor[3], it seems reasonable to approach this system building project with a Pritt stick, rather than Loctite. In other words, the toolkit should allow the designer to start small and simple, with a complimentary experimental approach, where new technical solutions and concepts can be unstuck, recomposed, corrected, or reapplied where needed. This contrasts with the oft bemoaned inevitable leaning towards creating expensive, inflexible, and linearly developed ecosystems of desired solutions against projected cost-savings and scheduling, or static concept documents and contracts that struggle to adapt to a changing environment.

We might trace the origins of this rigidity to the Industrial Era, where militaries organised under a modernist construct sought to ‘synchronize and reliably produce conforming behaviors and effects’[4] throughout the operational design and decision-making process.

Design Educator at USSOCOM, Ben Zweibelson, similarly locates Defence’s undynamic approach to system or operational design in its ‘analytic-linear worldview’, which emerges as an ‘inability to see alternative ways of strategizing because of the seductive nature of analytic and linear constructs for military organizations’. Far be it from me to argue against the seductive nature of analytic and linear constructs… Nevertheless, such a worldview could prove fatal to achieving MDI and its underpinning C4ISTAR system given its highly ‘wicked’[5] nature.

The Berghain Problem aka Defence's Techno-Fetishism.

The complexity challenge for Defence professionals is further compounded by the absolute centrality of technological innovation to the MDI concept. The ‘sense, understand, and orchestrate’ axiom is built on the assumption that integrating new, innovative technologies will reduce our uncertainty and allow us to seize initiative in the fog of a multi-domain operational environment. As the JCN 1/20 explains:

The information exchange requirement associated with MDI is unprecedented. The mass of data derived from myriad sensors will necessitate artificial intelligence and machine learning to detect patterns where previously there was only noise.

Described as ‘technical rationalism’ by academic Antoine Bousquet, this example represents a specific technology-centric strategic approach to reducing or gaining order over complexity through an overarching functionalist paradigm[6]. We see this in UK Defence discourse and organisational structure, too. Indeed, the commercial arm charged with driving pan-Defence MDI initiatives and related transformational projects is aptly named the ‘Future Capability Group’. This may just seem like semantic nit-picking, but such a title does perhaps offer another example of Defence’s preoccupation with technology, rates of change, and the promise of future capability as opposed to, let's say, understanding how we rapidly integrate existing and legacy systems, which can also help to drive a more bottom-up approach.

Such preoccupation with tech innovation is based in what has been described by some as presentism - the tendency to find the current era to be exceptionally, even uniquely, turbulent, and past eras to seem tranquil in comparison. Or is it a case of neophilia, the belief that what is observed and experienced in the battlespace is entirely novel. Indeed, this is hardly a new mode of thinking for Defence professionals. Take this quote from Bernard Brodie’s pretty zany 1949 journal article, ‘Strategy as a Science’:

The modern officer accountable for strategic planning and decisions has a burden of which his counterpart of a century or more ago was quite free. Nelson could spend his lifetime learning and perfecting the art of the admiral without any need to fear that the fundamental postulates of that art would change under his feet.[7]

The point isn't that new technology won't be necessary to building such a system, nor that we shouldn't pursue many of the groundbreaking capabilities that will surely save lives in the future. Rather, I merely highlight that a future capability festish, and assumptions about future capabilities in multi-domain operations, could hinder the MDI learning and experimentation adventure, as well as distract from the more important (but maybe less sexy) issues of organisational and cultural modernisation that will enable desired innovations to occur.

Let's talk about what we think we know and how we've come to know it... or something like that.

In his 2011 paper on System Design and Project Management Principles, US procurement expert Mark Mandeles also highlighted the limits of our understanding when it comes to designing complex systems and integrating new technologies:

Like it or not, information and knowledge about military capabilities are limited and imperfect. To deal with this situation, a process is needed through which knowledge is communicated, acquired, and applied. Consequently, the solution to the problem of organizing the acquisition process is to harness and direct the interactions of people and companies—each of whom possess, more or less, only partial knowledge about the task at hand.

Here, Mandeles references the epistemological and ontological issues inherent in strategic planning and decision-making in Defence. That is, the problem of assumed knowledge about how technologies will develop and become operationalised, assumed environmental and market stability and predictability, and the institutionalised “normal science” of capability development which we see translated into the very organisational structures that contain it i.e. a lack of alignment between the ‘people and companies’ involved, stove-piped systems, and programme offices.

To address this, new approaches to acquisition have been championed by the MOD. This includes scaling their Category Management procurement approach, engagement with high-tempo experimentation and exercise programmes, use of capability Centres of Expertise that draw upon the knowledge of industry, academia, operators, Dstl, and Defence Digital, and the creation of an ‘innovation bridge’ to bring these Rusholme Ruffians together. There’s also the standing up of various public-private engagement spaces, such as the latest MDI Working Group, the Royal Navy’s Maritime Enterprise Planning Groups, and the Army’s BattleLab, to generate closer alignment across Defence and fill in gaps in market knowledge.

'In an increasingly complex world... We've got to innovate!' said everybody ever.

That said, it’s not yet quite clear to me how these initiatives will translate into the most important and tangible outcomes – actually realising this new system of technical solutions and getting them into the field. Maybe I’m a little numb from years of repeated rhetoric on innovation. Lauren Schousboe writes:

Innovation, transformation, and revolutions in military affairs are conditions you actively and purposefully create. If the discourse remains monotonous and bland, it becomes harder to translate intentions into action on the specific technologies and challenges. Though the feeling of urgency is constant, the official rhetoric on technology and military change should reflect the strategic and technological reality and not come in general terms… When the official discourse remains unspecific and unchanged, it becomes vulnerable.

Whilst working groups and other kinds of hubs for knowledge sharing remain important in this quest to drive future innovation, we should also put as much emphasis on understanding and overcoming the structural barriers to implementation (safe passage through the Valley of Death) and defining the core metrics of success, whether that’s lead time, deployment frequency or, most importantly, The One Metric That Matters (OMTM)’, as they call it in Kessel Run – the value that any new solution brings the user, or the combat capability that is being delivered. Further, whilst new innovation bridges and centres of excellence and hubs and whatever else present viable access points for small and non-traditional defence businesses, it's imperative that the support is there to ensure they are being connected into the appropriate customers and users across Defence.

Indeed, we should heed the warnings from Tim Grayson, DARPA’s Strategic Technology Office lead, who discussed the challenge they faced supporting the DoD’s JADC2 development efforts in an interview a couple of weeks ago:

“We’ve got a portfolio of about 20 technology programs spanning different functional areas providing the tools needed to do this very rapid, just-in-time integration to meet mission needs. Our real big challenge now is how to transition those. We don’t have a nice, clean-cut program office where we can take our tools and put them directly in a classic, conventional program of record. They’re not designed to do those kinds of things. They’re enabling infrastructure.”

Relax, rewind, reframe.

In their report, also published earlier this month, Lt Gen David Deptula and Heather Penney of the Mitchell Institute of Aerospace Power Studies discussed how undynamic organisational structures and processes hinder rapid technology insertion in the U.S. Air Force. They recommend the creation of a ‘Systems Program Office’ that would be ‘dedicated to developing these [Mission Integration] software capabilities, funding these tools independently of traditional weapon systems, and creating specialized funding architectures that can keep pace with software development’.

Ironically, then, we see a response to the DoD’s ‘bureaucratic and funding barriers’ overcome by the creation of more bureaucracy. As opposed to, say, the destruction of arcane processes or outdated policies and behaviours that do not enable the kind of fast adaptation and learning needed.

Perhaps an alternative approach, as referenced in the MDI JCN, might be to start ‘reframing’ MDI procurement and technology development. In a report by MITRE, the authors argued that ‘portfolios should be the foundational structure for requirements, budgets, and acquisitions’. As such, they outlined this idea of Mission Area Portfolios (MAPs) that would allow much greater flexibility and adaptability within a capability portfolio. They write:

What if the key performance parameters in a requirements document removed thresholds and only had objectives? Programs would continue to measure and report on cost, schedule, and performance, yet acquisition and operational portfolio managers would have tradespace flexibility. If an 80% solution can be delivered sooner and resources shifted to higher portfolio priorities, they could do so. If inserting a new technology increases costs and schedule, but increases mission impact, the program isn’t penalized. Removing some of the requirements and acquisition constraints would enable MAPs to be more responsive to changes in operations, threats, technologies, budgets, and system performance to maximize impact.

This approach has been expanded upon further by those at DAPRA working on the Mosaic Warfare concept[8]. The authors write:

To enable decision-centric warfare, a portion of the defense budget should be aligned around important missions, such as defeating air defenses, strike, or ASW. Capabilities and experimentation related to these mission areas could be aligned by placing them together under the appropriate portfolio. This mission-centred budget could also fund the development of communications linkages and system interfaces to improve interoperability.

A similar approach in the UK could also drive much faster and more effective capability development. Cross-functional, pan-governement groups could be organised around organising ‘recombining known or soon-to-be fielded technologies’ (think Pritt Stick, not superglue) so that they can experiment with varying compositions of technologies, people, and processes around a common mission goal, and perhaps create something innovative along the way (Minecraft was born out of combining Rollercoast Tycoon, Infinimer, and Dwarf Fortress, after all!). Mission-oriented programme offices and campaigns of experimentation might also help to shift us away from technical rationalism and ‘stovepiped’ platform- or domain-specific capability development, and encourage traditional defence industry to become more systems-oriented and mission value-focused.

Furthermore, as recent reports from the US remind us, the difficulties in realising new joint operating concepts should not be underestimated. Technological solutions, particularly informational ones, should be explored, yes, but not at the expense of addressing the cognitive, psychological, and organisational challenges that peer adversaries present too.[9]

Certainly, as JCN 1/20 acknowledges, the cultural change necessary to achieve MDI puts ‘a much-increased demand on professional military education and is at least as important as any other capability requirement described’.

[1] JCN 1/20, p. 69.

[2] JCN 1/20, p. 31.

[3] Other craft-adjacent metaphors in Defence include ‘knitting’ or ‘stitching’ together sensors, and data ‘fabrics’.

[4] See Taylor, Lt Col Brian J (2017) Intelligence and Design: Th­inking about Operational Art.

[5] On ‘tame and wicked problems’ see Rittel, HWJ. & Webber, MM. (1973) Dilemmas in a General Theory of Planning. Policy Sciences (4).

[6] See Zweibelson, Ben (2021) On American military strategic frames and self-imposed limitations: designing disruptive and alternative constructs.

[7] Brodie, Bernard (1949) Strategy as a Science. World Politics (4). 470-471.

[8] The Mosaic Warfare operating paradigm goes beyond the Systems of Systems model towards a more bottom-up approach where large numbers of smaller, disaggregated individual (majority autonomous) elements form different composable ‘force packages’ before and during operations.

[9] See Dougherty, Chris (2021) More than Half the Battle Information and Command in a New American Way of War. In the paper, Dougherty challenges the notion of ‘Information Superiority’ and proposes a new paradigm, that of ‘Degradation Dominance’.