Futures Lab - Cutting edge technologies and research influencing Defence

Graedon Crouch, Futures Lab Command Lead for Air, looks at the latest technology news that can help shape Defence thinking.

The ‘Palletrone’

Researchers from South Korea have developed an aerial cargo platform called the ‘Palletrone’ – a portmanteau of pallet and drone. Cargo is placed on top of the drone which maintains a hover using its inertial measurement unit. Users can gently push the hovering Palletrone which will then move in that direction, much like pushing a shopping trolley without the effort.

As a proof of concept (initial tests were conducted with 3kg of cargo) the Palletrone is unlikely to be sufficiently mature for battlefield use. In the future, it may find a use for moving equipment around uneven and difficult terrain.

Reducing air-conditioning energy use

The Economist suggests that air-conditioning is responsible for more CO2 emissions than the aviation industry, and with rising global temperatures, the use of air-conditioning is only likely to increase. Conventional air-conditioners remove moisture from the air as well as cooling it in an energy intensive process. Moisture absorbing desiccants are a more energy-efficient method, and new metal-organic frameworks can dehumidify air using up to 40% less energy in trial machines. Other new approaches innovate on the century’s old technique of water evaporation for cooling, removing the need for widely used greenhouse gas refrigerants which can leak.

Austere military deployments can require temperature-controlled environments such as field hospitals and IT facilities. Achieving this with less energy consumption will reduce the burden on fuel supply logistic chains.

Wireless connectivity through concrete

WaveCore has developed an ethernet bridge capable of achieving Gbps connections through thick concrete walls. Using a 6Ghz signal and directional antennae, initial testing achieved 4Gbps through 30cm of concrete.

The ability to quickly install networks in buildings of opportunity used by deployed forces can be a challenge. Commercialisation and adoption of this technology may accelerate their initial connectivity.

Battery developments

Over several decades, battery cathodes have changed considerably: nickel, sodium, zinc, etc. but lithium has (so far) taken the crown. To improve battery performance, experimentation is now turning to other parts of the battery. Anodes are typically graphite, but silicon can increase energy density by up to 50%. New electrolytes, including solid-state, are also being trialled to improve performance in high and low temperatures. And the electrical current collectors – typically aluminium and copper – are being redesigned for improved performance.

Grid-scale batteries store energy when supply is high (typically from renewables) and release it when demand is higher than supply. The cost of grid scale storage is falling and the adoption of chemistries such as sodium-ion and nickel-hydrogen (more suited to large, static installations) is accelerating grid plans for additional battery storage capacity.

The global environments in which the military operate are often harsh, with extreme temperatures, vibration, and shock. Constant improvement in battery resilience and performance will improve military utility and adoption beyond trials and demonstrations.

Robotic control models

Traditionally, AI models controlling robots need retraining on new data or tweaking if they are to carry out a task in a different environment. Robot Utility Models (RUMs) developed by New York University, Meta and Hello Robot sidestep this need. In new environments, tasks such as opening drawers and picking up items were successfully completed 90% of the time without retraining.

Similarly, researchers from the University of California, Berkeley have developed a universal robotic control policy that can be applied to different types of robots. A relative paucity of training data, combined with a multitude of robot embodiments (how many arms, wheels, propellers, etc. the robot has combined with their spatial relationship and freedoms of movement) make this a significant challenge. Their CrossFormer model uses a transformer architecture which matched, though not exceed, previous performance. Also, the model is too large to run on typical robots and must be run on an external server and is only just fast enough to keep up with real-world application.

The adoption of battlefield robots remains dogged by challenging terrain, environments and changing circumstances. These advances in robot control and operation may incrementally improve the performance of future battlefield robots.

Follow Futures Lab on LinkedIn to keep up to date with our latest news.