By Robbin Laird
In his presentation to the Williams Foundation seminar held on September 28, 2022, the head of Force Design in the Australian Department of Defence, MAJGEN Anthony Rawlins, addressed directly the question of how the ADF could realistically and effectively ramp up its capabilities in the mid-term. This is how he put it: “Has the hardening of expensive, exquisite, arguably irreplaceable platforms now reached its logical zenith? This is manifest in the arguments for the cheap or the expendable as a supplement or potentially a replacement for expensive crewed platforms going forward.
“Defence is not just investing in exponential developments in autonomy, artificial intelligence, remote sensing, etc, etc as an R and D line of effort. But defence is doing so with a view to fielding capability in the immediate short term. And it hardly meets the definition of survivability to be investing in platforms and capabilities that are designed to be expendable.”
And at the Williams Foundation seminar on Next Generation Autonomous Systems, April 8, 2021, Professor Jason Scholz, CEO of the Trusted Autonomous Systems Defence Cooperation Centre, put the challenge and opportunity for the ADF along similar lines. Here is what he emphasized during his presentation to the seminar: “Autonomous systems for air, land, sea, space, cyber, electromagnetic, and information environments offers huge potential to enhance Australia’s critical and scarce manned platforms and soldiers, and realizing this now and into the future requires leadership in defense, in industry, in science and technology and academia with an ambition and an appetite for risk in effecting high-impact and disruptive change.”
Scholz sees autonomous systems as providing mass to the distributed force. This is what he noted in my interview with him conducted after his seminar presentation: “Humans express mission command goals to machines, machines express to the operator what actions they can take to achieve that, and a contract agreement is formed. Within the commander’s intent, machines then subcontract to other machines and so on, dynamically adapting as the battle evolves to build that Mosaic.” In both his presentation and our discussion, he highlighted a capability on which they are working now that can provide for sensors and communications capabilities to support the force which complements manned assets to provide for
In both his presentation and our discussion, he highlighted a capability on which they are working now that can provide for sensors and communications capabilities to support the force which complements manned assets to provide for Information, Reconnaissance, and Intelligence. In other words, autonomous systems can provide for sensor networks, which can be part of the effort to leverage information systems to deliver more timely and effective decisions.
The ADF is already undergoing a transition to shape a distributed integrated force. Next Generation Autonomous Systems can provide a further set of capabilities for a more effective, dense, survivable, and capable ADF as it builds out for operations in the Indo-Pacific region and enhances its defense of the Australian continent.
How software-driven payload to mission systems can accelerate ADF capabilities in the near to mid-term was already highlighted in the work conducted earlier by the Plan Jericho team in their focus on the importance of being able to gain transient software advantage in conflict with adversaries. Where we are headed is in a direction which could yield significant operational advantages whereby code re-writing is driven by operations and operations by training, and training driving development and looping back again into operations.
The graphic below was generated by the Plan Jericho team early in their efforts and captures what transient software advantage means, and is a key capability which can be delivered by the systems which the head of force design suggests are critical to ramping up ADF capabilities in the near to mid-term.
I had a chance to discuss the way ahead with regard to autonomous systems built around rapid software development delivering payloads to missions with the head of Anduril Asia Pacific during my September 2022 visit to Australia.
David Goodrich OAM is the Executive Chairman and Chief Executive Officer, Anduril Asia Pacific. Before Anduril, David advised the Australian Defence Force (ADF) and other Australian and State government entities on dozens of major weapons system acquisition, sustainment and infrastructure contracts (his biography can be viewed at the end of this article along with three elements of the current Anduril approach in defence systems).
(his biography can be viewed at the end of this article along with three elements of the current Anduril approach in defence systems).
Anduril came to my attention after my meeting with the head of the Royal Australian Navy’s autonomous systems programs, Commodore Darron Kavanagh. In that interview, Kavanagh underscored the following: “As soon as I say I’ve got requirements for a combat system, I immediately go into a classical system engineering approach. But that approach doesn’t actually allow for the agility necessary rapidly to change that combat system.
“If I look at classical primes, they are often hardware first companies, software second. And there’s a lot of legacy in the design.
“One of the things we’ve been looking at is how would you take a software first approach to accelerate our maritime autonomous systems capabilities. This is one of the reasons that the sovereign industry players that we’ve selected recently to work with in the autonomous systems areas are software driven in their development rather than platform focused.”
Specifically, Kavanagh and I discussed the partnership between Anduril, Defence Science and Technology Group (DSTG) and the Royal Australian Navy to build extra-large autonomous undersea vehicles. This is how an article by Edwina Callus published by the Australian Department of Defence published on June 1, 2022 described the partnership:
Defence has partnered with global tech company Anduril to design and develop extra-large autonomous undersea vehicles (XLAUVs).
Under the co-funded arrangement, Defence scientists, Navy personnel and Anduril robotics specialists will produce three prototype XLAUVs over three years, delivering a manufacture-ready vehicle at the end of 2025.
XLAUVs are typically between 10 and 30 metres long and have the capacity to carry various military payloads over long distances.
In addition, a large diameter autonomous undersea vehicle will be delivered to Australia as a testbed vehicle to enable experimentation, testing and validation to commence from the first quarter of next year.
It is expected that extensive industry engagement and ongoing initiatives to broaden Australia’s uncrewed undersea vehicle industry will ensure Australia is at the forefront of robotic autonomous systems.
Chief Defence Scientist Professor Tanya Monro said the collaboration was significant, not only for the critical capability it would deliver to Defence, but because it demonstrated how innovative new technology could, should, and would be fast-tracked and streamlined to keep pace with a rapidly changing strategic environment.
“We have long recognised in Defence that we need to transition innovative concepts into capability more quickly. That urgency to deliver impact is what shapes our Defence innovation programs, and is the driving force behind More, Together, our 10-year Defence science and technology strategy,” Professor Monro said.
“Our new approach to innovation is all about focus, scale and impact.
“By partnering shoulder to shoulder with our industry and Navy colleagues to co-develop this critical capability, that is exactly what we’ll achieve.”
The ambitious XLAUV development program will establish the foundations of an Australian sovereign XLAUV capability while strengthening Defence’s understanding of the technology associated with operating these platforms.
The project is partly funded by Defence’s Next Generation Technologies Fund.
“This is a great example of our innovation system in action and exactly the sort of activity that the Next Generation Technologies Fund was designed to enable,’ Professor Monro said.
Head Navy Capability Rear Admiral Peter Quinn said the program would focus on the capability first, and use technology to find solutions to problems.
“We will build a little, test a little and learn a lot,” Rear Admiral Quinn said.
For Anduril Australia, the XLAUV program is the first of many significant investments in Australian industrial capabilities designed to deliver disruptive autonomous capabilities to the ADF and allies.
Executive chairman and CEO of Anduril Australia, David Goodrich, said that through this important partnership, Anduril Australia would become a major player in the thriving defence industrial base in Australia and contribute to Australia becoming a leading exporter of cutting-edge autonomous defence capability to the rest of the world.
The XLAUV program is directly aligned with Defence’s strategic research activity, the Remote Undersea Surveillance STaR Shot, a mission-directed research program aimed at securing Australia’s maritime interests through the provision of persistent and responsive undersea domain awareness.
They have the potential to provide the ADF with an important, stealthy, multi-role, undersea capability, complementing and enhancing the agility and potency of the Navy’s current submarine and surface combatant force in maintaining peace and stability in the Indo-Pacific region.”
What does it mean to be a software first defence company?
Anduril was created by its founder, Palmer Luckey, as a software driven defence company. This is not surprising as Luckey made his fortune by inventing and then commercialising a new generation of virtual reality headset goggles called the Oculus Rift. Software development was at the heart of his ability to shape a new generation of low-cost virtual reality headset googles that then generated rapid growth in the gaming market and in adjacent markets where virtual reality has become a driver for innovation and change.
After he sold his Oculus company to Facebook (now Meta) for several billion dollars, Luckey focused on building a new type of defense company, one driven by software development and the ability to integrate sensor data into the kinds of common operating pictures crucial to the decision superiority which our militaries have set as a key goal for combat dominance. As Goodrich highlighted about Luckey: “he’s always been innovation-driven and had a software first mentality.”
After selling his Oculus company, Luckey focused on the national security market as he felt it was ripe for a new approach and one driven by a software first mentality. This is how Goodrich put it: “Palmer wanted to invest into something associated with national security because he was passionate about it. He looked at all of the defense primes and couldn’t see them adopting a software first mentality or acting at consumer product speed. He didn’t want to put his money into a company that adopted yesterday’s kind of defence industrial approach as he knew this wouldn’t win the next war.
“So he decided that the only way he was going to have an impact on national security in the way that he wanted to was to pull together a group of like-minded individuals and start a company which was a software-first security and defence company. That company is Anduril.”
Goodrich then laid out the path followed by Anduril – which has now led to the partnership with the Commonwealth to build XLAUVs. “The founders of Anduril initially saw an opportunity to disrupt and innovate in the border security area in the United States. They worked hard and came up with a software driven technological solution using artificial intelligence, edge computing, and computer vision to create a capability to scan vast areas of the southern border region in the US and be able to automatically categorize via artificial intelligence what was being scanned to precisely identify where an event or item of interest was unfolding and then relay that information on “one pane of glass” using Anduril’s custom operating system called Lattice.
“LatticeOS is a bespoke operating system which fuses together a limitless number of sensors to filter out the important information from the unimportant information and then display that information on one screen which we refer to as a single pane of glass.
“And everything Anduril has done since this time has utilized the same Lattice software platform. Lattice is what really makes Anduril different. We are a first and foremost a defence software company with approximately two thirds of our almost 1,400 people software engineers and only one third hardware engineers. What we’re about is using machine learning, autonomy, edge computing, computer vision and artificial intelligence to continuously update and evolve our Lattice operating system.
“We operate in an open-architecture environment where we have actually integrated more third-party sensors into our Lattice operating system than Anduril sensors. Data from a diversity of sensors can be integrated seamlessly into Lattice and be displayed and filtered and utilized to give the war fighter data that they need to make better strategic decisions.
“From the very beginning, our company has been engineered to be different from a what a legacy defence prime contractor looks like today. We look more like Google, or Amazon or SpaceX or Atlassian than a defence prime contractor like a Lockheed Martin or a Raytheon or a Boeing or a Northrop Grumman.”
And the approach to an ongoing focus on writing code and reshaping the software resident in platforms is driven by working with the combat forces, to know what they need and what they would like to see evolved. Driven by ongoing combat demands, a software driven defence approach can provide for the kind of agile software advantage highlighted by the Plan Jericho team.
This is how Goodrich describes Anduril’s working approach with the warfighters. “We partner with the people who are on the front line and understand the threats. 20% of our employees are veterans who understand the key issues facing the warfighters, having lived that role for most of their careers.
“However we have a mantra in the company that says, “we always take very seriously our defense customer’s understanding of the critical problems they are confronted with every day, however they don’t always have a solid understanding of the range of potential solutions to these problems.”
“Our defense customers don’t have the necessary range of people possessing the range of tech experience or the breadth of knowledge required to understand what the best possible solution is”. To adopt their thinking towards the best, most creative solution is ultimately not in the service of the agile game-changing outcome they deserve and are looking for. This is where Anduril comes in.”
What does software transient advantage mean for the evolution of combat capabilities?
We then discussed Anduril’s engagement in Ukraine. I will not provide excessive details here for obvious reasons, but I will describe the process. Anduril has a UAV called the Ghost.
This is how Anduril describes the Ghost:
“With real-time intelligence, surveillance and multi-mission reconnaissance capabilities, up to 55 minutes of continuous flight and with near-silent acoustics, Ghost is mission ready.”
It is powered by the Lattice Operating System. “Lattice brings indispensable aviation capability without increasing complexity and cognitive load for operators. And it enables point and click mission planning. “Plan missions, flight operations, task sensors and navigate all with the click of a button, enabling safe and effective flight with minimal instruction and training.”
It is an expeditionary system. “Ghost is a man-portable system that can be assembled and readied for flight by a single operator in less than 3 minutes. Ghost is waterproof, weatherproof, and multi-mission capable.”
It is a VTOL system. “Ghost’s single-rotor design enables vertical takeoff and precise landing in confined spaces.”
With a kill web concept of operations, the focus is upon the payload to mission and not the platform as the core focus. “Ghost is designed for easy payload integration and field maintenance. Networked payload bays and quick-release gimbal allow quick module change in the field and safe handling of sensitive components.”
Anduril on their web page provides quotes from users which highlight what Ghost is all about. For example, a UK Royal Marines, 40 Commando is quoted as saying: “What’s different with Ghost is that it’s built for soldiering purposes. It’s always searching, it’s constantly looking. This isn’t just a drone with a camera, it’s AI.”
With the Lattice Operating System at the core of the Ghost mission system, what does rapid software insertion then mean?
According to Goodrich: “We’ve evolved five versions of the Ghost product in five years. We’ve created four versions of the Sentry towers in five years. We are iterating across the entire software stack and product suite to evolve and reshape the capabilities of Ghost, as a platform reflecting the software innovation and hardware capability.”
He then focused on an ADF example. “We have sold Ghost to the Australian special operations command (SOCOMD). And we’re working with this leading customer to rapidly adapt their Ghosts to better meet their evolving CONOPs. The SOCOMD team saw opportunities to tailor the user interface (UI) to be more aligned to their particular use. In response to the customer feedback, we re-engineered the user interface in less than eight weeks. We are about to deploy with them again with the new UI that is exactly what they wanted. It is this kind of rapid turn that we’re able to deliver into an ADF environment.
“Although this kind of approach is not appropriate for every kind of cornerstone defence platform, the obvious need to increase the number of autonomous systems that are partnering with and leveraging the power of our exquisite and extremely expensive systems, to defend our peace and prosperity, and if called upon to win the next war, means that the agility modelled by Anduril needs to become business-as-usual.”
How then did Anduril build the platform which hosts the software?
As Goodrich highlighted: “We don’t vertically integrate for vertical integration’s sake. We’d much prefer to horizontally integrate. Take the case of the Ghost UAV.
“A UAV is not a new or novel concept. We tried initially buy one. That’s what we tried to do, however what was on the shelf couldn’t do what we needed it to do. Nothing met the concept of operations we knew our customers needed.
Now to Ukraine. It is public knowledge that the Anduril has been supporting the Ukraine response. And without disclosing too many details, it is fair to say that our technology has been very effective. We have done live rewrites of the software code to respond to Russian counter-measures in real time. Continuous software updates essentially refreshes our transient advantage on the battle-field. Rapid iteration is the signature of a software first defence prime of the future.
As Goodrich underscored: “whilst it’s absolutely devastating to see what’s going on in the Ukraine at the moment, it does enable the West to understand the speed and agility that is needed to be able to operate if we are going to win and close the kill web.”
Looking Ahead to the XL-AUV
We then discussed the XL-AUV. Obviously, the model for shaping the capability will follow in many ways, the Ghost model as already described using the Lattice OS as the core operating system.
Goodrich explained the way ahead with the program as follows:
“We’re developing three prototypes in three years. Subject to these prototypes meeting the RAN’s capability need and we will be delivering a manufacturing plan to roll XLAUVs out at significant scale. The three prototypes themselves will be rapidly evolving over the contact period using the principle of build a bit, test a bit, learn a lot, then repeat. The team will be constantly be upgrading the subsystems, however the platform itself is going to be mature and reliable.
“With regard to power systems, our focus is upon duration. Our endurance will be significant. We are building with the goal of the platform being able to operate for weeks if not months before it needs a recharge or a refuel. They will be able to go to sleep on the ocean floor for long periods.
“Subject to government approval, we will be able to manufacture the XLAUVs at scale in a few years’ time. And what does it scale mean? We could be talking 20, 50, 100 boats, meaning that the opportunity and challenge for the RAN will be to shape a concept of operations as to how to deploy this incredibly disruptive and powerful capability.”
He concluded: “I am proud to lead such an incredibly talented and driven team here in Australia. Anduril has turned the traditional defence industry on its head at a time when it is needed. Working with the innovation leaders in the Royal Australian Navy (RAN) and DSTG is a great privilege for all of us at Anduril Australia – the excitement and energy in our office every day is makes it an amazing place to work. It has been a great step for me personally after a decade of commercial advisory work the Department of Defence.”David Goodrich OAM Anduril BIO