Inside Army's Innovation Engine: LTCOL Adam Hepworth PT 1

In Part 2, Adam Hepworth gets onto the technology he finds most interesting. He makes the case that maturing uncrewed systems with emergent behaviour into something genuinely deployable could break the one-to-many problem and deliver an order-of-magnitude jump in both capability and capacity. In conversation with host Josh Bailey, Adam talks about brain-controlled robot dogs and the human-machine interface being the real unsolved problem. He talks about RICO working on exploring what is the next groundbreaking technology, and the workforce challenge underpinning it: how does Army grow the AI operators, maintainers and support skills that the Army will need. The episode closes on responsible AI, drawing on Adam's work with the Global Commission on Responsible AI in the Military and its Responsible by Design report, and his conviction that the entire point of Army's RAS strategy is to protect people first. Key takeaways Swarms as tipping point (28:03 to 29:23) The Oppenheimer of 2030s (29:39 to 30:54) The workforce problem (31:59 to 33:34) Train RAS like driving (33:34 to 35:38) The interface is unsolved (35:38 to 37:35) Brain-controlled robot dog (37:35 to 39:36) Protecting people first (39:36 to 43:57) If you missed Part 1 on Army's innovation engine and uncrewed battle runs, start there.

This week Josh sits down with LTCOL Adam Hepworth, former director of RICO, the Australian Army's Innovation and Experimentation office. Adam walks through the breadth of what RICO covers, from robotics and autonomous systems and AI to quantum sensing and battlefield electrification, and how the office pairs emerging tech with operational concepts and then puts it in soldiers' hands to test. Adam describes RICO's last large-scale autonomy activity - a live-fire armoured squadron battle run with crewed and uncrewed units fighting together. His main point is that the technology is rarely the hard part. The real work is the enabling infrastructure around it, the policy, the legal reviews, the training and the test and evaluation that turn a one-off demonstration into a deployable, sustainable capability. Key takeaways What RICO actually does (00:14 to 03:30) Human Machine Interfaces (05:03 to 08:28) Human Machine Teaming - battle run (09:26 to 12:42) The hard part (12:42 to 14:29) Soldier feedback (15:01 to 17:21) Adams Journey (18:25 to 27:53) Stick around for Part 2 next week!

In Part 2, Emily Hughes returns to talk risk with Josh: specifically, why we hold machines to a higher standard than the humans they replace, even though we know most accidents stem from operator error. It's a sharp lens on the regulatory caution that's slowing down uncrewed systems. In conversation with host Josh Bailey, Emily digs into the case for right-sizing risk appetite, the difference between dropping a five-kilo drone in an empty paddock and putting an aircraft into a crowd, and why small companies are essential precisely because they'll experiment in the grey, riskier spaces that larger organisations tie themselves in knots over.  Key takeaways The pufferfish: aircraft or munition, and why layering every regulation is the wrong instinct (19:08 to 21:23) Leveraging Simulation: run it a thousand times, then fly the fifty that matter (21:23 to 23:33) We demand more of machines, despite knowing where the errors come from (23:33 to 24:17) Why small industry is essential in the riskier spaces that the bigger players avoid (24:17 to 26:06) Stitching things together: sovereign players, quantum tech and a venture capital lens (26:06 to 28:07) Can boys engineer too? The remarks that keep us grounded (28:07 to 30:54) Optimism about the Defence Delivery Agency and learning from the past without repeating it (30:54 to 34:14) If you missed Part 1 on human-machine teaming and calming the pufferfish, start there.

This week Emily Hughes joins Josh to talk about the real work of making autonomous systems useful: not the technology itself, but the humans who have to trust it, operate it, and integrate it into how they fight. An aerosystems engineer with around 25 years across air and land programs, Emily has spent two-thirds of her career in autonomy, from a crewed gyrocopter she instrumented back when all of this was still a science experiment. Emily makes the case that UAS is finally becoming business as usual, one tool available to a commander rather than a separate squadron off in the future. She credits the ADF's experimental units for de-risking that shift, and argues the remaining hurdles are less about technology and more about regulation, compliance and a culture she affectionately calls the pufferfish reaction to risk. The conversation also covers her advice to small companies trying to break into defence, why speed is their real advantage, and how to find a way in through established players rather than fighting the process alone. Key takeaways From space-obsessed kid to aerosystems engineer, and why solving problems beats plane-spotting (01:37 to 03:03) Technology for its own sake helps no one: building systems around how the operator actually works (03:03 to 04:59) Harriers, jet fuel and a terribly unsafe gyrocopter: hands-on autonomy nearly 20 years ago (04:59 to 07:31) UAS as business as usual: one tool for the commander, not a separate squadron (07:31 to 10:42) Calming the pufferfish: why the real hurdles are now regulatory, not technical (10:42 to 13:31) Hidden complexity under a simple surface, and why intuitive systems mean the hard work happened in industry (13:31 to 14:53) Breaking into defence: find momentum, partner with established players, let your agility count (14:53 to 19:08) Stick around for Part 2 next week!

In Part 2, Alex Bristol makes the case for getting physically closer to the threat, and he's putting his money where his mouth is, heading to the Baltics and Eastern Europe to learn from the Finns, the Poles and Ukraine, because that's where the ground truth is. In conversation with host Josh Bailey, we get into the gray-zone use of uncrewed systems as a way to test a nation's will without triggering a hard response, the blurring of civil and military lines, and the constraints on counter-UAS in the homeland where kinetic, EW and laser options are largely off the table. Alex frames the geopolitical shift bluntly: power projection is back, and adversaries are probing how far they can push. The episode closes on something more personal: Unit Decode, the veteran-transition platform Alex is building, and his conviction that using AI well, being able to do the job of five to ten people, is the edge that matters coming out of service. Key takeaways Getting closer to the threat: why Alex is heading to the Baltics and Eastern Europe (15:08 to 17:14) Uncrewed systems as a gray-zone probe: testing will without a hard response (17:14 to 18:39) Civil-military blurring, and Chinese vessels fitted with missiles and drones (18:39 to 19:28) Hands tied at home: no kinetics, no EW, no lasers on the civilian front (19:28 to 21:44) Power projection is back, and the fight for firm specs on attritable systems (21:44 to 25:45) Unit Decode: rebuilding veteran transition around voice-based AI coaching (25:45 to 29:48) You can't be afraid of AI: it's a tool, not the Terminator (29:48 to 31:25) If you missed Part 1 on counter-UAS ground truth from Syria and why complexity kills, start there.

This week Alex Bristol joins Josh to bring something a bit different to the show: ground truth. A former US Army Special Forces Green Beret, Alex left service, felt the pull back, and last year found himself in Syria working Counter-UAS on a rocket-based project, standing on the ground as a defense contractor with, in his words, nothing in his hand. In conversation with host Josh Bailey, Alex makes a blunt case that US defense technology has built itself too complex, too costly and too slow, when the soldiers who actually have to use it, often National Guard or conscripts with limited train-up, need solutions that are stupid simple. He argues a 50 cal with a CROWS mount and an infrared sight can knock down a Shahed, and that a $73,000 pair of drones just doesn't scale. It's a hard look at the gap between what industry ships and what works when a high-threat window hits and everyone's order goes out the window. The conversation also covers how the Syrian drone threat compares to the saturation in Ukraine, and the shift toward smaller, attritable, swappable-sensor systems that units can actually afford to lose. Key takeaways From 18X-ray off the street to Green Beret, and the pull back to Syria (00:50 to 03:05) Ground truth on a rocket-based counter-UAS project, and kinetics beyond drone-on-drone (03:05 to 05:46) Syria versus Ukraine: why the threat depends on the actor and their wartime economy (05:46 to 07:40) Too complex, too costly, too slow: the case against exquisite US systems (07:40 to 09:26) When the threat window hits, order goes out: simple kit you can grab with your eyes closed (09:26 to 11:57) Attritable and swappable: a 50 cal, an hourglass, and why $73k a pair doesn't scale (11:57 to 15:08) Stick around for Part 2 next week!

This week Darren DelSignore returns for Part 2 of his conversation with Josh, focusing on the counter-UAS response now taking shape across the West and where DroneSec sees the next frontier of drone intelligence. Darren walks through Joint Interagency Task Force 401, the August proclamation that has driven a major uptick in US, NATO and Latin American counter-UAS activity. He details what good intelligence sharing now looks like across allies, why the world's largest stolen drone database matters (including drones stolen from US military bases as recently as last week), and how the details matter in Threat analysis - an agricultural drone seen over downtown Sydney can tell you more than a generic detection. The conversation also looks at where DroneSec is heading next, including interactive red-versus-blue simulations built into its TTP database, and why humans, not machines, remain the alpha in a counter-UAS fight. Key takeaways (00:00 to 02:30) Collective intel beats adversary speed (02:30 to 05:00) JIATF 401 mobilises the West (05:00 to 07:30) Red versus blue at scale (07:30 to 09:30) The cyber 2016 inflection point (09:30 to 12:30) Interactive TTPs and live simulations (12:30 to 15:00) Drones stolen from US bases (15:00 to 17:00) Ag drones over downtown Sydney (17:00 to 18:00) Humans hold the alpha If you missed Part 1 on the cyber parallel and the new criminal toolkit, start there.

This week Darren DelSignore joins Josh for Part 1 of a conversation on the parallels between cybersecurity and UAS threat intelligence, and why drone-related risk is reaching a decade-long inflection point. Darren is the General Manager of DroneSec, joining the company after five years at Google running sales for VirusTotal, the cyber threat intelligence platform widely used across the public sector, military and intelligence community. Darren maps the current UAS threat picture onto the early days of cyber: a constantly shifting cat-and-mouse game where today's most effective counter tool can be obsolete in a year. He argues the intelligence layer matters more than any single piece of hardware, and explains how a $1,000 drone and a few hours of practice are already letting organised criminals move contraband into prisons and even lift people out of them. The conversation also covers why bureaucracy gives bad actors a real speed advantage, and why a globally shared intelligence picture is what gives the good guys a chance to keep up. Key takeaways (00:00 to 02:30) From VirusTotal to DroneSec (02:30 to 05:30) Finance, then cybersecurity, now drones (05:30 to 09:00) Why the intel layer wins (09:00 to 11:00) Today's tool, tomorrow's relic (11:00 to 13:30) $1,000 buys you a problem (13:30 to 15:00) Bad guys move faster Stick around for Part 2 next week!

This week Joni Sytsma returns for Part 2 to talk about the Western drone industry's inflection point, building counter-UAS interceptors from drone components, and her vision for sovereign Australian manufacturing. Joni argues that the US ban on foreign drone components is creating a vacuum Australian makers can fill, provided they can build the picks and shovels: the autopilots, flight controllers, motors and ESCs that every platform depends on. She explains her own approach at Rocket Fast Drones, where in-house fabrication is cutting costs and reducing reliance on overseas suppliers, while a recent test flew one of her autopilots to 5,000 feet on a rocket in New South Wales. The episode also looks at her safety-first destructive testing program, why she believes smart mass delivered by rocket-launched interceptors could be a missing piece of the counter-UAS puzzle, and her vision for keeping Australia worth living in. Key takeaways (01:01 to 03:30) US bans and the Western inflection (03:30 to 07:00) Picks and shovels, not just gold (07:00 to 09:00) Big Bird and SpaceX dreams (09:00 to 11:00) Safety, thermals, walking the drone (11:00 to 13:30) Aerospace jobs for Brisbane (13:30 to 17:00) Rocket-launched counter-drone interceptors (17:00 to 19:30) Keeping Australia good If you missed Part 1, go back and catch the build-it, crash-it, rebuild-it story first.

Wing Commander Keirin Joyce returns for Part 2 of his Tactics & Tech conversation, turning to where uncrewed systems are heading next. Drawing on his PhD in technology prediction, Keirin argues the hardware problems are largely a matter of time, while the real challenges lie in software, autonomy, and trust. In conversation with hosts Andrew Crowe and Josh Bailey, Keirin explains why certification has become the long pole in the tent for the whole industry, from AI-generated code that 1990s-era standards were never built to handle, to hydrogen-electric propulsion no aviation safety authority yet knows how to certify. The tension is clear: in a society that grants no trust until something is certified, regulation may be a bigger brake on progress than technology. The episode also looks at how military drone racing builds social licence for drones, opens STEM pathways into defence industry and academia, and creates adaptive sport for injured and veteran personnel. Key takeaways (00:00 to 01:15) The ScanEagle began as a tuna-spotting aircraft before soldiers repurposed it for surveillance, a perfect example of Keirin's theme that technology becomes powerful or not depending on how people use it. (02:28 to 04:55) Military drone racing is gaining momentum, with upcoming events in the US and UK, growing interest from Indonesia, the Philippines and Singapore, and a possible event in Australia's north around 2027. (05:22 to 06:45) Hardware is advancing steadily, but software, AI and autonomy are the hard problems, along with how to trust code a human did not write. (06:52 to 07:55) Certification is the biggest bottleneck, with software standards rooted in 1990s aircraft practice and poorly suited to AI-generated code. (08:23 to 09:55) Hydrogen-electric propulsion is a major focus for new designs, yet no safety authority anywhere has a regulation for certifying it. (09:55 to 10:55) Regulation, not technology, may be the primary constraint, with the real challenge being the balance between under and over regulation. (12:36 to 13:25) Australia has grown to around 30,000 remote pilot licences in a decade, while the regulator's capacity has not kept pace. (13:25 to 16:25) Military drone racing is building social licence, STEM and recruitment pathways, and adaptive sport for veterans, with a pitch for Invictus around 2031.