Improbable Defence

We need open, accessible content to make collaboration our competitive edge

Collective training blog

 Mike Raker headshot

Mike Raker, Chief Technology Officer

We’re feeling the effects of the Fourth Industrial Revolution (4IR) everywhere and all the time. Our jobs, businesses and lifestyles are being transformed by 4IR technology. Self-driving cars are part of a not-too-distant future thanks to advances in robotics and AI. An ever-evolving Internet of Things gave us 10+ billion smart devices in 2020. Cloud and edge computing and other innovations kept us connected even as COVID-19 kept us apart – the list goes on.

A hallmark of the 4IR is the use and reuse of technologies, such as AI and machine learning, to produce unprecedented technological innovation. Since these technologies feed on data, it’s a good thing that said data has never been more prolific and accessible than it is today, though we still have a ways to go in the national security sector.

With the rapid rise of the internet came data silos  decentralised, fragmented stores of data across an organisation that remained isolated, often when departments or businesses chose to compete instead of working together toward common business goals.

In the national security sector, data silos arose due to security concerns, acquisition approaches, and a portion of the defence industrial base’s desire to maintain rights to data as a competitive edge  an advantage to the company, but not the mission.

In response, governments started to mandate more rights to data, resulting in broader access. This led to larger repositories of raw and curated data from different sources that is accessible by a multitude of different end users and services. In turn, this facilitated increased use of that data for AI/ML and other applications. This corpus of results, inclusive of data-driven models and the industry’s production of new capabilities like digital twins, creates the concept of ‘content’.

Where data is the ‘oil’ of the 4IR, content is the gold but its value as such isn’t being realised due to content silos that make it hard to integrate, expensive to develop, and slow to adapt.

While we’ve advanced from data to applications and content, the latter are typically built for a singular purpose, and this is what creates a content silo. For example, you can use data to generate a model of a communication network, but is that model extensible and flexible enough to interact with adjacent content such as models of the producers and consumers on that network? If not, it’s isolated, stovepiped content. Break it open and you can achieve results that are many times more powerful than the original application as a true flywheel of value.

Siloed content impedes the use and usability of the content itself

To take agent-based models as an example, the value of a model of a population is that it serves some purpose. But it’s more interesting and useful when multiple models behave together in the same world, in a rational and expected way. The increase in value of coupled models comes from them being more than the sum of their parts.

To realise this value, we need to make content accessible in the same way we did data. But applying the same techniques of how we solved data stovepipes to content stovepipes won’t work.

We solved the siloed data problem by implementing services and standards which exposed data to external consumers, thereby creating data lakes so two systems or end users can easily exchange information. With content, it’s far more complex. Models need to talk to each other, and when these conversations happen, we need to be sure that each model acts in a reasonable way in response to the data on which they’re built, driving the need far beyond data and interface standards.

Virtual world experimentation for real-world application

This is especially true when we consider synthetic environments (SEs). These provide a safe virtual proving ground to create and test all sorts of ideas, from policy choices about the size and shape of the Armed Forces, to designs of new systems and technologies and exploring how best to use them.

Crucially, SEs are a critical capability that make use of the latest, most reliable and most relevant content to help users explore scenarios in a virtual world before taking action in the real world. It’s therefore essential that this content whether it’s models, systems or systems of systems  interoperates and integrates so as to make it as valuable and effective as possible within the SE.

We need to break down content silos if we’re to make progress both within the field of synthetics and as a society. Industry needs to find ways of working together to ensure that this content becomes ever more accessible to allow policy makers and decision takers to rapidly understand and adapt to the fluid, fast-moving challenges of tomorrow.

The good news: we’re getting there. In this 8-part blog series, we’ll showcase the power of partner collaboration in solving some of the most intractable problems this ideal faces.

We’ll explore examples of where we’ve worked with industry to break down content silos, accelerate the composition of content to achieve a mission outcome, and many other novel concepts in this field. In so doing, we aim to support the creation of synthetic environments that help governments understand the world as it is, explore the world as it could be, and so operate effectively in complex, fluid and often highly ambiguous situations.

Visit our Newsroom for more news, insights and opinions spanning the defence sector.