Gaining military advantage with 5G

“We shall get a big pipe with 5G.  And there’s no doubt we shall fill it.”

However, the greatest concern expressed by one of the panellists at SPADE this week is how they put the right things on it.  SPADE is an annual defence and security event.  This year it is being held at Soestduinen in the Netherlands, sponsored by IBM, Samsung and others.

One high priority domestic opportunity that 5G affords is for disaster relief because it helps the citizen most.  Responders can optimise logistics to get equipment in and insurers resolve claims quickly.  5G offers high network bandwidth with low latency.  Its emergence is coupled with increasing storage capacity at the edge.  Furthermore, analytics and AI are part of this ecosystem to deliver quality and timely information as situations unfold.  5G is also here today, with the first operational network running in South Korea.

Opportunities for defence

In a military context, 5G offers data supremacy on the battlefield.  All the information a commander needs becomes accessible on a mobile device powered by 5G.  Drones and vehicles can extend the network from the physical endpoint into the battlefield much more cheaply than using today’s dedicated and more limited radio communications.  Such networks are set up ad hoc allowing much more operational flexibility at the tactical edge.  In addition, each endpoint both receives and transmits.  This offers greater resilience than is possible today, and an ability to create local ‘supercomputers’ by linking devices without needing to connect back.  It speeds up local decision making and changes how we fight.

5G is the enabler for integrating AI.  Data can be processed and analysed at the edge, but network bandwidth also allows data to be brought back for searching.  The proportion of data captured by sensors that can be exploited goes up.

This combination of 5G and AI presents greater insight at the edge.  Common operating pictures are become richer, more relevant and current.  For example, targeting data is accessible faster to make a decision and act before the target has gone.

Furthermore, there is the chance to undertake command and control differently.  Placing more decision making power is in the hands of the platoon leader reduces the need for so many headquarters.  Spending could be redirected in favour of more combat soldiers.  For example, 6-7 people support each person in combat in the US Department of Defense today.  Could that ratio be halved?

Evolving security

The attack surface will become larger because of greater and wider use of technology with 5G.  Growth in the number of devices and the rise of the Internet of Things increases perimeter so security will need to be approached from a perspective of zero trust.  However, 5G overcomes 4G’s omni-directional limitation and its variable bands make it hard to jam the network.

Our current approach to security is based on the notion of defending the centre.  We try to make mobiles as secure as the centre with its perimeter.  However, 5G is designed around the end user, and perimeters become ad hoc with pop up 5G networks.  Consequently, there is a shift away from classic network architectures and security designs.  This presents challenges today that need tackling to realise the true value of 5G.

One example is who gets access to the networks?   Assured identity challenges will need tackling as networks become fluid, meshed and disconnected because we have centralised identity services today.  Phones carrying our identity may be part of the answer.

It is no longer good enough to secure networks and endpoints rather than the entire system.  5G allows a single security system without an operating system.  This may well change cyber economics because today adversaries apply most focus to exploiting the operating system, and the defender has a greater opportunity at lower cost to disrupt the attack.

The need for separate military-only networks might disappear with 5G, no longer cost effective.  They would span 99% of a geographic area for a small subscriber base if Government did choose to allocate dedicated spectrum.  On the other hand, a telco has a large subscriber base and focussed on covering 99% of population.  Furthermore, an ability to use consumer devices in the military drives down cost.

It is possible that the benefits of 5G will accrue to businesses and governments first rather than consumers and this will drive rollout.  One example could be rural areas to deliver a better emergency response.

Re-training the workforce

Defence departments face challenges today with the time it takes to get authority to operate – achieving compliance to put an application on the network.  This will need a change the mindset by the custodians of network security to one that understands the evolving threat and applies risk management.  Change management is necessary to bring in new security architecture.

Access to a labour force is a growing problem with an ageing population in many developed countries.  People are the biggest cost driver and 5G offers opportunities for savings.  Redistributing and re-skilling the workforce are options.  Border control could become agentless at the gates with image recognition powered by AI on 5G, with lower centralised human support.  Network specialists could be re-deployed to the edge as network and cyber experts on the battlefield, recognising that two battles may now need to be fought in the first mile.

5G is rolling out today and the opportunities are limited only by our imagination.  But there are challenges for defence departments that need to be tackled.  Training for 5G will take time.  We need to start now.

 

How ready is the public sector for AI?

Artificial intelligence (AI) should now be seen as a core part of business transformation rather than merely an interesting technical project.  It means revisiting the relationship between government and its citizens, and rethinking how public services are delivered.

There are already numerous applications deployed that use AI in public sector.  These fall into five areas:

  • Improve customer service contact centres. Assistants are being used to increase both civil servant and citizen satisfaction through greater productivity and accuracy, and extended hours of support.  There is a reduction in mundane work for civil servants, and the burden from citizens on specialists is reduced.  Benefits are being realized within a few weeks.
  • Enhance knowledge workers. AI is particularly attractive in fields with massive volumes of domain-specific data to find patterns that offer improved results.  Fields include legal and regulatory, policy development, oncology, cyber security, and more generally taking this approach helps those on rotation become productive more quickly.
  • Manage the complexity of risk and Contract governance is one such use of AI by Governments.  Operational decision making has also been augmented by AI by monitoring current situations, assessing risks and making recommendations.
  • Find the best talent and modernise learning. AI is being used to analyse the talent market to find candidates who best fit a role.  Aptitude can be assessed to help build digital skills in scarce areas, eg cyber security.  Furthermore, AI aids learning in content, its delivery and management.
  • Empower developers to build AI-powered Equipping business teams to build applications using AI tooling and training platforms has facilitated integration of AI with both existing systems and emerging technologies such as blockchain.  Governments have been able to provide access to video content for its citizens using audio analysis, improve the way they deliver services using speech to text, and better protect critical infrastructure.

The opportunity apparent here is to deliver better services to citizens more quickly whilst reducing the burden on civil servants.  General characteristics can be drawn from these implementations that can be applied to assess those processes that are suitable for AI and likely to deliver benefits.  These are:

  • Does the process exploit a lot of data? And could it benefit from using other accessible content?
  • Is a personalised service required?
  • Is the process repetitive and reliant on a degree of knowledge and intelligence?

Business teams will need to be prepared to move away from the way things have always been done.  One technique to imagine new possibilities with AI is to creatively explore problems from an end user perspective using Design Thinking.

AI needs IA

Too few artificial intelligence (AI) projects succeed.  Many organisations approach AI believing that you can collect data for an algorithm in the hope that it realises the anticipated benefits.  Instead you should look at data and design a system to address a problem, not an algorithm.

Here are some keys to success for adopting AI.

  • Select the right business problem. This must be one for which a team already exists and has the data.  It avoids the pitfall where, “We need to test AI,” results in a deceptively attractive initiative which has low business value and is hard.  For example, a business process is required to collect data.  Nevertheless, there is a conundrum for many organisations that the business case to get the data requires a demonstration of AI.
  • Look at the data. Typically, organisations significantly underestimate the effort needed to orchestrate the data in readiness for AI.  AI needs accurate data, and data cleansing and preparation takes 80% of the effort.  This is a hard engineering problem and requires a sound approach to information architecture, technologies and a range of skills, not just data scientists.
  • Build systems, not algorithms. Many assume that a sequence of steps is sufficient to generate insight and recommendations.  However, feedback is crucial to improving overall accuracy.  It is complex with lots of moving parts and demands a multi-disciplinary approach.

AI must be transparent for the public to trust it.  This is especially significant for the public sector because important decisions must be explainable.  It is essential to understand who trains the AI system, what data was used to train it, and what went into the recommendations made by the algorithm.  This extends the realm of information governance.

In summary, AI needs IA: Information Architecture.