How organisations can leverage the full power of data
Frontier technologies such as artificial intelligence (AI), machine learning (ML), quantum, and high-performance computing are radically altering the national security landscape.
Organisations need to understand and harness these shifts if they are to master this disruption, enable effective change and deliver better outcomes.
In this new series from Shephard Studio, we explore what happens when technology challenges us to develop new ways of doing things.
In examining today’s technological landscape, we also consider Japanese cultural practices and concepts from which innovators can draw inspiration.
Kaizen is a Japanese approach to creating continuous improvement based on the idea that small, ongoing positive changes can reap significant improvements.
Governments, national security organisations and companies across the defence sector today have access to a huge array of data. Gathering this resource is challenging enough – exploiting it is another matter entirely.
The importance of data in today’s world has implications for how governments prioritise and invest in specific technologies.
Not only is the volume of available information increasing exponentially, but the need to understand the data’s origin, including what other data sources were used to create it, and its veracity is essential.
‘Not only do you have to keep pace and try to make sense of that data, and increasingly increasing volume of data,’ explains Richard Carter, a computer scientist and strategic advisor on AI.
‘But you also need to understand and try and figure out how to sift the wheat from the chaff – what’s real versus what’s fake? What should we be paying attention to?’
Here, technology may also provide the answer, with machine learning and artificial intelligence processes able to 'triage' the data and present only the information of genuine interest to human analysts.
Such technologies offer an opportunity to realise the full value of data, according to Joanne Benbrook, a Senior Solution Architect at Fujitsu Defence and National Security.
‘At the moment, the information tends to exist in silos, and it’s challenging to bring it together and make it cohesive,’ Benbrook explains. ‘There’s a great opportunity to look at how we can better integrate that information.'
For example, this could involve more sophisticated aggregation and analysis of open-source intelligence – such as video newsfeeds and social media – with information and applications that are held centrally.
‘Some of that data is quite large – we don’t necessarily want to bring it into the centre,’ Benbrook says. ‘But we might want to consider whether we can process it at the edge, how we develop applications that can access that data at the edge.’
The capacity to leverage open-source information has been apparent throughout the war in Ukraine, where the availability of commercial satellite imagery is proving crucial, explains August Cole.
Cole is an author who looks to the future of conflict through fiction, an approach known as fictional intelligence or FICINT storytelling.
‘The enabling technology is not so much the hardware that’s orbiting, but the AI systems that are sifting and sorting, using image recognition to figure out what’s important,’ Cole says.
While this is only a snapshot of the role technology is now playing, it is illustrative of a shifting landscape.
The future information environment will be increasingly multimodal, with a need to work across platforms and teams in a consolidated manner to avoid the silo problem.
That’s the belief of Dr Darminder Ghataoura, Director of AI and Data Science at Fujitsu Defence and National Security.
Ghataoura points to the move towards data integration, notably a knowledge-driven approach based on a kind of 'data dictionary' or common language across different teams.
‘It’s not just about bringing data into a common place… how can I explore that data and look for the specific relationships and contexts? How can I connect the data even further by extracting those relationships and connecting concepts into a kind of knowledge graph?’
Such data challenges can be broken into different stages.
Firstly, there is data capture – understanding what sensors do. Data capture includes passive cameras and other physical sensors, along with combing the internet and similar activities.
Next comes management and storage, which brings challenges around the volume of data involved.
‘It becomes increasingly important to filter at this stage, because you cannot get the bandwidth needed to get all the possible sensor data back to where you want it to be. You need to filter what you want at the edge,’ outlines Dr Dave Snelling, Director of Advanced Compute at the Fujitsu Center for Cognitive and Advanced Technologies.
Snelling explains that the third phase is understanding, which includes several different aspects. One is to know what you are looking at, understanding its structure and the relationships between pieces of information that may have been independently captured. It is also important to consider data augmentation in the understanding category.
‘We take information that we have from multiple different sources and recognise that there is a cognitive connection that we’ve made based on either human cognition or AI cognition,’ Snelling explains. ‘There’s more there than then meets the eye. You make those connections, feed those connections back in.'
The final step involves drilling down to the essence of the information within data through decision-making.
‘You’ve got all this information – now what do you do?’ Snelling poses.
As Richard Carter explains, this is where AI processes act as a 'teammate' to a human analyst by filtering vast amounts of data and highlighting the salient information. The analyst can then spend their valuable time interpreting the data.
‘You need a human in the loop working with the data and turning that into actionable intelligence. So, the question becomes how AI can assist the intelligence analyst rather than replace the analyst,’ Carter argues.
One area where this already paying dividends is the development of AI-fuelled predictive maintenance solutions.
For instance, one initiative under development is the integration of sensor data with maintenance data. This improvement allows military services and other organisations to become more agile in maintenance, moving from a fixed maintenance schedule to one driven by the needs of a particular platform.
‘The idea behind the approach was to distil the key anomalies within that dataset,’ Ghataoura explains, ‘looking at patterns which could diverge from normal behaviour.
‘Let’s say you had an engine temperature reading coming through… the AI algorithm understands whether [that matches a] normal operation.’
Such approaches help a manager focus their time, resources and costs.
‘If you could do that far earlier in the time horizon, you can get that spare part in time for the service to be carried out. It gave the customer the ability to understand what value they can exploit from the data they have.’
Quantum computing is one technology area that will provide a step change in speed and scale for handling computational problems.
There are many more algorithms and capabilities to be discovered as we learn more about quantum computing, says Ellen Devereux, Quantum Computing Consultant at Fujitsu UK.
As organisations grapple with ever-increasing volumes and complexities of data, quantum computing will one day provide the means to analyse and utilise this vital resource to a degree not possible today.
‘The bigger the problem, the more difficult it is to model with conventional compute,’ says Devereux.
‘Quantum computing will be able to solve some problems that are currently intractable with conventional computing and speed up the processing of other problems.'
Fujitsu is already making connected progress through its Digital Annealer, a high-speed computer that provides a bridge between today’s systems and the quantum computing of the future.
The sum of these technological parts presents a considerable opportunity for national security organisations to draw on. August Cole goes one step further and argues that the flood of data that’s now available opens an ‘aperture to see the world differently’.
‘From a human cognition point of view, how machines “see” the world is crucial to understand and figure out how to communicate back to humans what they’re seeing.
Kaizen is a Japanese approach to creating continuous improvement.
Formed from the words Kai (change) and Zen (good), the concept is based on the idea that small, ongoing positive changes can reap significant improvements.
Typically, it is based on cooperation and commitment and stands in contrast to approaches that use radical or top-down changes to achieve transformation.
