Reflecting on Last Year’s Conference
The inaugural Disruptive Technology for Defence Transformation conference took place in London on 24-25 October 2017 and examined how disruptive technology will transform defense capability and operations. It was attended by: senior military from many countries; NATO: technology firms such as Google and QinetiQ; leading aerospace and defense companies including Northrop Grumman and BAE Systems; research institutions such as IISS and OSTL; and academia including Cranfield University and the US National Defense University. The Conference was sponsored by Northrop Grumman Europe and chaired by General Sir Richard Barrons.
The Conference examined the current and anticipated operational environment and existing approaches to force modernization, setting the scene for exploring the wide span of disruptive technologies that by and large are being led by the commercial sector. The core issue for the Conference was how these technologies, whether singly or in combination, may be adapted or adopted to transform defense capabilities and operating methods in a way that delivers extraordinary advantage over larger, more traditional armed forces.
The Conference considered that is very likely that much of this transformation can be accomplished at lower cost than present programmes given the logarithmic cost reductions exhibited by information-based technology. The Conference touched on the spectrum of requirements of Counter-Terrorism and Insurgency, Hybrid confrontation, and conventional conflict at scale, pace and great lethality between peer armed forces. Urgency behind this work was widely acknowledged, in the face of a more challenging, less stable multipolar world in which both state conflict and the threat from resilient non-state actors are escalating risks. There is already a deficit against some potential adversaries in cyber, domination of the Electro-Magnetic Spectrum, Air Defence, and ability to maintain access to space-based capability. The easy global availability of a great deal of new technology means that the process of 21st century transformation in order to protect homelands and vital interests is a race already underway at pace. This transformation will not easily be conducted in secret unlike previous military transformations.
The Conference considered the challenges of acquiring disruptive technology. It is clear that the organisation and process used to acquire large conventional platforms such as ships will not work for the provision of fast-paced technology where two years is regarded as the longest tenable horizon for anticipating what technology will deliver. It will be necessary to change how specifications are written, allowing room for constant development, and to invest in experimentation in the knowledge that some work will fail. To introduce disruptive technology into defence requires an acquisition system that emulates how technology companies act in a highly competitive environment.
The Conference concluded that bringing together international military, government, alliance, technology, academic and industrial representatives to establish relationships and understanding in this arena was essential and should be repeated annually.
Conclusions Drawn From 2017
Data will power most transformative change, starting with how intelligence is collected, fused, analysed and presented to leaders and users. The faster ‘big data’ and machine learning is drawn into the heart of intelligence, surveillance, and target acquisition capability to replace traditional methods the quicker a decisive edge can be established. Conversely, the importance of stealth in all environments and the requirement to conceal battlefield signatures becomes a core requirement.
Military Command and Control will become faster, more accurate, more agile and resilient the faster it can introduce machine learning supported by robust data and communication services. This has the potential to significantly reduce the size and vulnerability of deployed HQs, and the potential for reach-back to resilient static national and alliance HQs. It suggests a major saving in staff manpower is achievable.
Many militaries, in all domains, are actively experimenting with the potential for a manned/unmanned mix of capability. In particular, the combination of data, connectivity, robotics and autonomy looks to be the route to restoring mass. resilience, deception, and combat power when compared to the cost of conventional platforms and volunteer forces. It will not be long before some Services ‘bake in’ manpower cost savings on the back of robotics.
There will be a profoundly difficult debate about the limits on autonomous weapons, with clear advantages in some uses, such as the defence of major installations and platforms from attack at speed and scale, and great risk in others, especially where lethal force without a ‘man on the loop’ may be applied around civilians.
There is no doubt that Virtual and Augmented Reality will become pivotal to seamless individual and collective training, experimentation, concept development, planning and mission support. The technology will permit training that cannot be done in the physical world. The drive is for single synthetic environments that support all these activities in one solution.
There are several emerging technologies that require more concentrated attention and experimentation as they have demonstrably clear potential to disrupt conventional military structures and methods: synthetic biology, the use of graphene, and 30 printing for operations and support were all examined. In each case the civilian lead requires clear, focused military engagement.
Round Table Discussions
LETHAL AUTONOMY: CONCERNS & CHALLENGES
Led by Colonel Dan M. Sullivan, Chief of Stoff MCWL, Deputy Director, Futures Directorate, U.S. Marine Corps
When discussing AI it is important to d distinguish between neural networks and Algorithmic Based networks. AI is a spectrum of capability that is evolving rapidly. It will be critical to define terms precisely going forward (e.g. partially autonomous, fully autonomous, man-in-the-loop, positive identification). There is a certain amount of unpredictability of an entity that is vastly more intelligent than its operator. If AI proves to be many times more accurate than humans at determining PID, would it be ethical to restrict the machine from making “shoot/don’t shoot” decisions.
Current indications are that our adversaries are pursuing this capability and the development of AI is not occurring in a controlled environment like we had during the first and second offsets. Education and transparency with the public is critical to generate support for whichever path policy makers choose to pursue regarding lethal autonomy.
Recommendations: The recommendations were that signing any international prohibitions regarding lethal autonomy at this time would not be wise in order to preserve decision space for policy makers until we more fully understand this capability. There should be continued interdisciplinary research on this topic in order to make informed decisions regarding lethal autonomy going forward. In the meantime, nations should issue a “no first use” policy regarding lethal, autonomous systems.
AI OPPORIUNIDES & THREATS
Led by Paul Winstanley, Former Executive Director Innovations, UKDSC, Independent Consultant
There is a potential EW / cyber threat for AI, for example, an AI algorithm could be “fed” inappropriate training data to reduce its initial effectiveness. Building trust between the AI solution and the warfighter was a key discussion point. In particular there is a need to understand, as after action reporting, why the AI made a particular decision. Understanding the rationale would help build trust. AI could be deployed initially in training and simulation systems to start building the necessary trust. There is an interest in using AI as part of a war game, in particular as OPFOR, and being able to train the AI based on the potential adversary’s previous tactics, the same AI solution could then be used as an operational “coach”. There is a risk of being too reliant on commercial AI training and validation as the defence operating environment has some specific, and unique, attributes. AI is seen as being useful to manage the mass of data to inform a human decision but it is recognised that some situations don’t allow sufficient time for a human reaction. In this case, using AI for a non-lethal response would be easier than for a lethal response. Recommendations: It was proposed that AI is certified based on standards of performance rather than an absolute and objective test; essentially AI is assessed in the same way as humans. A roadmap for AI application would be beneficial i.e. which are the priority “easy” implementations to the most challenging.
BREAKING PARADIGMS AND IMAGINING POTENTIAL FUTURES
Led by Dr. Lydia Kostopoulos, College of Information and Cyberspace, National Defense University, U.S.A.
Informed decision making can be used to strategic effect across domains; both the battlefield and when furthering national security interests. Thorough risk assessment is required in efforts to monitor, assess, and give right to authorities to take the risk. There is a need to know who owns the risk and create a possible risk authority map. In order to overcome information overload strategic intelligence and machine learning can be use, in addition to human machine teaming. Networked learning can also be utilised, using people in conjunction with networked machines, devices and sensors. Looking forwards in terms of cooperation and communication, the paradigms are rapidly changing and so is the battlefield.
REFORMING ACQUISITION TO MATCH THE SPEED OF INNOVATION
Malcolm Warr, Member, UK Federation of Small Businesses
Taking into consideration that over 93% of private sector are employed in the UK work in SMEs there is currently too much linear process ‘lust to dust’ and there is a need to move towards circular procurement with shortcuts through lessons learned. More effective contracting is needed; small innovative acquisition needs higher individual procurement expertise than complex projects. The MoD needs to acquire an SME champion in order to insert new expertise at all levels and ‘gateway’ organisations are needed to improve communications and information exchanged. In terms of tangible benefits this can add real value and provide bespoke design within budget. Barriers to this as that the current system is too inflexible and there is lack of proven SME/Corporate contracting models. Other barriers include the blurred roles and accountability, holistic risk and poor time management. Corporates tend to dismiss value of disruptive technology and are designed to work with sustaining technologies.
VIRTUAL & AUGMENTED REALITY IN TRAINING, PLANNING, & OPERATIONS
Led by Air Marshal (retd.) Sir Christopher Harper, Former Director General International Military Staff, HQ NATO
The capability of AV/RV currently available can be as good as the “Viz” glasses depicted in Ghost Fleet and is already being used by TSA for facial recognition at US airports and by the Ukrainian Army. Potential opportunities for the training of operators of complex systems are obvious. The F35 pilot of the future will conduct only 50% of their training flying the real aircraft and the rest will take place in simulators. There are a number of F35 systems that cannot be used in peacetime. However, these systems require more than just pilot (or-pilot-to-pilot) training, and Intel Analysts, engineers, fighter and air traffic controllers will also need simulated/virtual training if their skills are to be honed. This begs the requirement for co-operative, distributed, networked training in a massive scale, virtual world. The limitations are that the realities of physical environment are challenging to simulate, for example bomb disposal training in an air conditioned VR training room vs the reality of a hot/humid climate when wearing a CBRN suit. It is also impossible to simulate or emulate real danger or threat to life. (Although pain and electric shock could be used if ethically acceptable). That being said, immersion in a virtual world does generate some sense of reality – including G-force, movement, balance etc. The practical utility of AR/VR technology is growing fast. It has significant potential for applications in the defence and security space but its capabilities are inadequately understood. Risks, benefits and opportunities need to be carefully considered and mapped.
SYNTHETIC BIOLOGY FOR MILITARILY USEFUL TECHNOLOGY
Led by Petra Oyston, Technical Fellow, Defence Science and Technology Laboratory, UK
The key issues facing emerging and disruptive technology include experimentation and the consideration of how to transition from civilian application to defence application. The discussion considered if the innovate UK was an appropriate mechanism when the defence sector cannot afford to do everything. Partnerships will underpin everything, there are too many problems to be considered and not enough funds to deal with them so competition must be avoided and be replaced by closer working and openness. Legislation will be unlikely to be able to keep up with the developments and targeted investment is required to develop further multidisciplinary communities and realise fully disruptive impacts of synthetic biology.
In order to establish support and an exploitation route for a capability that is very innovative, cross-cutting and might fit applications that “are not interested in biology” we need to calculate how low level technology can enhance high level capability requirements and also calculate a method to identify breakpoints or knees in the performance curve. It is important to identify what needs to be done specifically in military research that is not being done in the civilian arena, and also understanding that civilian exploitation is a necessary factor in making technology economically viable for military use. Identifying the factor that can add value to a civilian technology to make it valuable to the military tends to depend on individual’s innovative thought.
POTENTIAL OF 3D PRINTING FOR FUTURE DEFENCE CAPABILITY
Led by Michael Petch, Editor-in-chief, 3D Printing Industry
Throughout the Disruptive Technology for Defence Transformation event numerous speakers made reference to projects enabled by 3D printing. Perhaps notably these speakers did not always refer to additive manufacturing/3D printing directly, possibly this is indicative of how the technology has now become used widely enough to no longer warrant a specific name check.
Discussion of 3D printing’s potential can be characterised across three domains: machines, materials and software. In this first domain, the availability of Metal AM systems was illustrated vividly by the presence metal printer at the conference. With a price point approximately, half that of comparable systems the machine demonstrates the relentless march of progress and the growing accessibility of metal 3D printing.
Discussion around materials for 3D printing was addressed with topics including the use big data to rapidly develop new alloys and a reduction of dependence on scarce resources through the creation of meta-materials.
In the software domain issues around security of digital assets and topology optimization in line with biomimicry were considered as areas for investigation. 3D printing applications included the potential for economic disruption brought about via leapfrogging of under industrialized economies and the potential for AM to disrupt existing manufacturing strongholds – with the societal changes this may entail. Addressing mega-trends such as urbanisation and the increase in number of megacities, 3D printing was seen as one technology poised to find application – with initial examples provided in logistics.
THE POTENTIAL IMPACT OF GRAPHENE TECHNOLOGIES ON DEFENCE SYSTEMS
Led by Steven J Savage, Research director (materials technology), Swedish Defence Research Agency
Graphene and graphene-based technologies as we understand them today seem unlikely to be regarded as disruptive technologies, but will certainly be classed as sustaining innovation. Some defence applications would be very valuable, including: new, multispectral photodetectors for broadband sensors covering the visual and infrared, structural radar absorbers and structural batteries or supercapacitors. The latter would be valuable in emerging applications demanding extreme multifunctionality, such as small unmanned systems. Graphene enhanced lightweight ballistic protection may be practicable but this is a very complex application and it is too early to make any credible forecast.
Despite the immaturity of the field, thanks to the high level of civil interest some applications may be very easy to implement immediately, such as thermal interface materials to allow electronics to be more easily cooled. This would lead to longer lifetime, improved reliability and higher power output. It is likely that graphene-containing coatings and paints for improved corrosion resistance and anti-fouling will be developed and would lead to reduced maintenance costs. Graphene-containing coatings for electrical conductivity (lightning protection and de-icing) are being developed.
As with all new technologies, progressing from low TRL and surviving the valley of death is difficult. The forecast for graphene technologies is in this respect not too gloomy because several major companies are already investing heavily. This investment is aimed at civil products, so it is important to monitor, evaluate and adapt those technologies for defence requirements. A cost-effective way to do this is by developing communities of interest (small companies, larger companies, defence research agencies and academia). These should be informal and versatile, but nonetheless should do more than simple horizon scanning and technology watch. In addition, it is essential to evaluate the military usefulness of the emerging technology.
