Multi-tier unmanned systems begin to show their promise

The role of unmanned systems on the battlefield has increased exponentially since 2020, as the wars in Nagorno-Karabakh and later Ukraine demonstrated the utility of less complex systems to achieve intelligence, surveillance and reconnaissance (ISR) as well as strike support. The fielding of increasingly smaller and cheaper unmanned aerial vehicles (UAVs) greatly diminished the entry cost for militaries wanting to operate unmanned technologies, as increasingly smaller platforms could act as ISR force enablers and strike assets.

As the cost barriers have reduced, the ability to create and integrate subsystems – from sensors to munitions – has also changed. Reducing the requirements for expensive and bespoke integrations on simpler platforms has meant more systems have been deployed with a greater appetite for engineering risk on the integration of low-cost systems. These integrations have now supported greater system-on-system development, allowing for a greater devolution of surveillance and strike to a forward-deployed platform. As more weapons have been integrated onto a widening range of platforms, the sensor-to-shooter decision and effector chain has been tightened, supporting enhanced lethality and responsiveness for militaries around the world.

Moving to multi-domain, multi-tier

The ability to deploy unmanned systems from a mothership has been advancing, with experiments of air-dropped UAVs from high-speed aircraft being demonstrated in 2016 by the US Air Force. The shift to UAVs acting as motherships has been given impetus from operations in Ukraine, with the country’s Baba Yaga heavy bomber drones acting as carriers for multiple first-person view (FPV) UAVs to enhance their range.

As one-way effector (OWE) systems – both loitering munition types and modified FPV systems – have become prevalent on the battlefield in Ukraine, the time needed to respond to intelligence and threats has diminished. Reducing the sensor-to-shooter timeframe has been a fundamental requirement for obtaining an operational advantage and reducing an opponent’s force multipliers. However, the domains remain relatively the same: mothership platforms deploy an air-to-ground effector, and ground-launched FPVs remain as ground-to-ground weapons. Unlike sea systems, aerial FPV drones and many OWEs face limitations in their endurance due to their power supply and payload. As a result, patrol missions to identify targets need to be undertaken by supporting platforms that can have the endurance to loiter. From a mission perspective this can mean the sensor-to-shooter chain is lengthened, as a reconnaissance asset needs to identify a target and call in support that may take time to arrive, reducing the ability to act against targets of opportunity.

This mothership concept does, however, demonstrate a shift towards a multi-tier unmanned concept. Deploying smaller-class systems from a larger one can deliver greater force protection for commanders as personnel do not need to move into danger zones. Additionally, smaller subplatforms provide granularity of the intelligence picture to operators as they can get closer to a target while the mothership acts as a communications relay.

Ukraine achieved another first in the unmanned-unmanned teaming concept in July 2026, when an unmanned landing craft of the 1st Unmanned Systems Battalion of Ukraine’s 123rd Separate Territorial Defence Brigade landed an unmanned ground vehicle armed with a machinegun on the Russian-occupied Kinburn Spit, northwest of the Crimean Peninsula. This demonstrated that the two domains could interact and deliver effects from the maritime to the land domain.

Dividing between sensing and effecting

Pushing this operational concept into other domains, particularly maritime, delivers greater situational awareness for commanders, as well as the ability to surveil, strike, survive and resume.

The role of unmanned surface vessels (USVs) as more than an ISR asset has been brought into stark relief in the Black Sea. Ukrainian forces have conducted operations against Russian naval targets and interdicted ghost fleet ships. The operations further demonstrated how USVs as one-way effectors can operate in highly contested environments, as well as how their interception by vessels underway can be a difficult task.

Combining this new USV concept of operations with traditional tasking means that more capability is being demanded of unmanned platforms as enhanced sensor fleets needing to be able to support responses as quickly as they can detect them. Large unmanned underwater vessels (UUVs), such as the Anduril Ghost Shark extra-large autonomous underwater vehicle (XL-AUV), focus more on endurance and the provision of sensor feeds and less on effectors. However, size and costs mean that affordable mass is hard to achieve, leading to a requirement to find a middle ground for naval operations. The requirement for multiple USVs to conduct patrol, surveillance and strike missions means that there is a larger operational footprint going beyond a single ISR or strike mission and into a multi-level operation.

Rapidly developing and fielding new maritime systems

Within this operational need, the US Department of War’s Defense Innovation Unit recently issued a call for proposals for a mature, market-ready USV that could be integrated with a minimum of two unmanned aerial systems (UASs). The two subordinate UASs would need to be a mix of ISR and kinetic effectors to support autonomous ISR, strike and battle damage tasks, with one being recoverable and reusable.

The USV itself is required to have a minimum operational range of 200 nautical miles (370 km) and a sustained cruise speed of at least 10 knots (18.5 km/h), in addition to being able to operate in blue water, littoral and brown water environments. Critically, size is a major factor, with the vessel to fit inside a 40 ft equivalent ISO container without any specialised equipment. Submissions, due by mid-August, will require a fully integrated system for test and evaluation within 29 days of selection for testing later in 2026.

Operationally, the USV will also need to operate under denied, disrupted, intermittent and limited (DDIL) communications links, operate an open-architecture software model, integrate with third-party command-and-control software, and be able to autonomously operate, follow and intercept contacts of interest.

Integrating many subsystems

Integrating a host of subsystems into this requirement is not without its challenges. Drone-in-a-box concepts, such as those from Skydio and Atlas, support fixed, static and mobile container operating systems that can host, launch, recover and recharge a UAV. Combining this with a moving landing platform, such as those developed by Norway’s Stable, requires integration between onboard sensors to provide a stabilised platform for landing UAVs at sea, accounting for the pitch, roll and speed changes of a vessel to allow for fully autonomous recovery.

Numerous platform providers – such as Kraken Technology, Textron Systems, UFORCE and Saronic Technologies – all have established platforms that can be used in response to this requirement. However, meeting the form factor requirements with required hardware and software integrations may prove to be a complex task for USV builders looking to meet the needs of the call.

Pinning this overall concept together, though, is the role of artificial intelligence (AI). In requiring an autonomous system to operate with true autonomy, AI is needed to control the vessel under changing sea conditions, make decisions based on the mission parameters set down by a commander, and function within autonomy guardrails around the use of force. Creating an appropriate autonomous software capability in a short space of time requires either an off-the-shelf software platform, or an open-architecture system from an existing supplier that will facilitate rapid integration of a variety of subsystems.

Outlook

The technology call from DIU will focus on bringing a new capability to operational fielding in a short space of time, supported by the organisation’s demand for development at pace. Platform requirements, notably the size, will ensure that traditional shipbuilding concepts will be eschewed in favour of smaller, rapidly produced systems.

The integration of multiple unmanned systems onto the vessel, representing the next iteration of multi-tier unmanned systems deployment, will support a greater devolution of force from manned systems. Integrating multiple systems onto one platform will support further development of the multi-tier concept, de-risking the concept and promoting open-architecture systems.

The integration of AI technology into the platform will also support new concepts of operation and future capabilities. As development and deployment continues, so the trust in AI technology will be enhanced as user and operational experience grows.