How to Think About Next Generation Autonomous Systems in the Evolving Defense Landscape: The Perspective of Marcus Hellyer

By Robbin Laird

The recent Williams Foundation Seminar on Next Generation Autonomous Systems (NGAS) provided insights with regard to a number of key issues concerning their future roles. A key consideration is how will they enter the force, how will they be part of force transformation and how best to understand how the manned-unmanned force mix might evolve.

My own focus has been upon the evolution of U.S. and allied forces in a direction of shaping a distributed but integratable force able to operate in contested environments. And NGAS are clearly part of the future of how distributed forces will operate in the future. For example, the core change already seen with regard to the U.S. Navy and USMC is shaping ways to distribute the force more effectively and this evolving template clearly anticipates a growing role for NGAS.

The presentation by Marcus Hellyer, a well-known and regarded defence analyst at the Australian Strategic Policy Institute, provided a particularly insightful look at how one might consider the way ahead for the entry of NGAS into the force as well as how the overall force might be calibrated for the future fight.

I had a chance to discuss his presentation and his thoughts with him in a recent phone discussion and interview.   I started the discussion by summarizing how I read his presentation and approach. Hellyer looked at the broader topography of the forces and asked the question, where might autonomous systems fit into where we need to drive our force structure?  The force is moving in the direction of being more disaggregated and autonomous systems will be useful in that process.

How do we manage disaggregation and what might be the role of NGAS in that process as well as providing more cost effective options as well?

Question: Is that a fair way to summarize your argument?

Hellyer: “It is. I like you get a little frustrated when we bog down in endless debates about hypothetical futures for AI or the legal and moral and ethical aspects of autonomy. I’m not really after autonomy for its own sake. Autonomy is a tool, not an end in itself.

“Like you, I’m very interested in moving towards distributed concepts, whether you want to call them the distributed kill web or mosaic warfare or whatever the term is that you want to use, I think that’s where we need to be going.

“In part, it is also about cost effectiveness as well. To paraphrase recent US Navy CNO comments: Can we continue to build $3 billion ships to carry 32 VLS cells? We need to start harnessing the disruptive potential of disaggregated systems.

“To get out of that kind of cost death spiral, we need to disaggregate capability. The bottom line is simple things are cheaper. Simple things are faster to build, and so you can build a lot more of them. To do that, you have to disaggregate capabilities off three-billion-dollar ships into smaller entities. But to make those entities relevant, you then need to link them together so they can talk with each other and work together.

“But to me, you need to disaggregate simply so you can get more sensors and more weapons into the air or into the water, and to shape a more resilient system.

“In my presentation to the Williams Foundation I spent quite a bit of time looking at a kind of analogous transition. I did that to make a couple of points. The first one is that really disruptive transitions can sneak up on you quite quickly. If you’re not prepared for them, you will be taken by surprise. And we have a really good example of a complex system of systems using disaggregated components that is evolving in front of us. The example I used was Australia’s electricity sector. Our electricity sector has dramatically transformed just in a very small period of time.

“I highlighted a screen grab from an app that shows at any point in time where Australia’s electricity supply is coming from. A couple of weeks ago, we hit a point where over half of our electricity supply was coming from renewable sources. So that transition to renewables has actually progressed quite a long way. We’ve reached a tipping point. I know tipping point is a cliché but we’re there, not because renewables now provide all of our electricity or even most of it. That 50% point was obviously one moment in time when the wind was blowing, and the sun was shining. It’s not hitting close to 50% all the time. But last year, about 28% of our electricity came from renewables.

“We’ve reached a tipping point in the sense that the commercial sectors does not want to invest its own money in traditional forms of generation; no business wants to build a coal-fired power plant. No business really wants to build a natural gas power plant. Only the government is considering whether it will pay to do that. We’ve reached that kind of tipping point where business thinks that it’s not economically viable to keep investing in fossil fuel powered generation. A transition has occurred; it’s just our thinking hasn’t really caught up with it in some ways.

“And this analogy also speaks to the distributed issue with regards to defense. The new electricity grid is not based on a small number of very large generators. It’s based on a large number of generators that range from small to larger. Australia has one of the highest take ups of rooftop solar panels in the world. You have private houses around Australia who are now feeding electricity into our electricity grid. So we range from individual households up to very large generators. We still do have some coal-fired power plants, but out in the countryside, we now have very, very large commercial-scale solar and, as in America, very large wind farms.

“But that’s a different kind of grid. You’ve got lots of contributors switching on and off, feeding in at different times, turning on in response to the market, turning off when the wind doesn’t blow, or the sun doesn’t shine. It requires a much more flexible, agile kind of grid and requires some of the kinds of autonomous brainpower that a military distributed kill web would require.

“I talked about that because I think it’s a really nice example of a transition that’s occurring in front of our eyes, but it’s also a transition that I think has some really nice relevant analogies for the military in terms of a more distributed grid suggesting what force distribution can provide as well.”

Question: What you are describing is the emergence of a mixed power system, where clearly fossil fuel remains crucial in many ways and funding to generate such capabilities is crucial to Australia’s ability to have security of supply as well. It’s a different mix, but the mix is being driven by introducing new capabilities and new approaches. To your point, it also allows you to think about a more viable grid in terms of not being so dependent on a small number of larger generating plants and transmission belts.

In a way this is an analogy to a distributed military kill web where you are shaping multiple ways to generate the combat effect one wants rather than relying on single point of failure large systems. Is that a fair characterization of your argument?

Hellyer: “It is. I think one useful kind of observation from the electricity sector is that rather than focusing on the capability of the individual generator we need to lot at the resilience of the entire system. Renewables will always look poor compared to a really big traditional generator, whether it’s gas or coal or nuclear because a single facility generates less power But we need to look at their contribution to the resilience of the entire system. In that regard, a renewable grid or a combination grid is much more resilient in many ways.

“If you look at in the electricity sector, don’t look at the new technology as a like for like replacement for the old technology. We were talking earlier about the Osprey. If you look at the Osprey as simply a fast helicopter, you’re kind of missing the point.

“So in the electricity grid, large scale batteries are not really there to replace coal-fired power plants. If you look at them in that regard, they’ll look like very poor replacements. What batteries do is they can stabilize the grid because they can switch it on and off instantaneously in response to demand. And because they can switch on and off so quickly, they completely outperform traditional generators in response to the spot market in the electricity sector.

“They actually have a different kind of capability and one that it really out competes the traditional generators. They are smaller but to use a military term, they are much more agile. They play a different role, but a complimentary one with the larger, traditional power generators. The challenge when we’re looking at new technologies is not to look at them as like for like replacements of what we have now, but to actually see what they do differently–that seems to be a particular challenge for militaries where the continual, daily competition of the marketplace isn’t at work.

“Another useful analogy between the electricity sector and the military is the issue of sunk cost and how to avoid ‘stranded assets’. Whether we’re talking about an LHD or a coal-fired power plant, if it’s in service, we’ve put so much money into it that we’re going to keep using it so we’ll need to find ways to adapt it and find a way to use it usefully in that web or mix of technologies.

“If we use a private sector analogy for the military, what we need to focus upon is shaping a balanced portfolio. NGAS will enter the force to provide a balanced portfolio. And one of the potential advantages  with NGAS is an ability to put them into play much more rapidly than you can with large platforms, capabilities which enhance the force. They offer you the ability to keep your big, traditional investments relevant.

“The big question, and one we probably can’t answer yet, is how much money do you keep committing to build those exquisitely capable yet extremely expensive traditional platforms like frigates and submarines? When do you turn that off? And how much do you put into the newer autonomous technologies, which to some degree are riskier because we don’t know exactly what’s going to become of them. We don’t know exactly how to use them. We don’t know exactly how to integrate them together and how all the command-and-control networks are going to work.

“But as another speaker at the Williams Foundation conference said, we need to have ‘a bias toward action’. That is, we can’t solve all the issues around autonomy in the abstract. We need to experiment and invest and solve as we go. Again, the electricity sector offers a nice analogy. Initially we started out with a few people putting some solar panels on their roof. That only grew slowly at first and very few people thought then that they could drive our power grid, yet here we are today with nobody wanting to put new money into the old technology.”

Dr. Hellyer’s presentation in PDF format:

NGAS Hellyer. Presentation

Dr. Hellyer’s presentation in e-book format:

Also, see the following:

Next Generation Autonomous Systems: An Australian Perspective