Issue 55 Uncrewed Systems Technology Apr/May 2024 Sellafield’s UAV equipment l Applied EV Blanc Robot l Battery tech l Robotican’s Goshawk l UGVs l UAVHE RW1 rotary l Roboat UVD l Autopilots l Arkeocean UVD l UMEX 2024 l CycloTech UVD

Battery technology | Focus Another solid-state battery system that is currently sampling for evaluation uses a lithium-metal anode with a solid electrolyte. This has an energy density of 950 Wh/L for the active cell area, which translates into a specific energy of 260 Wh/kg for a cell. The specific power is 792 W/kg at a 3 C discharge rate. The cell has a 3 C sustained discharge for high-power applications, with a peak of 10 C with full recovery. Testing has shown a lifetime of more than 400 cycles to 80% capacity, which is key for UAV performance. 3D printing The move to solid electrolytes creates the opportunity to use additive manufacturing to print battery cells. It has the ability to quickly modify performance and create cells in non-standard formats. This is particularly important as the improved safety of the materials allows for conformal batteries that make up part of the structure of a vehicle. This can reduce the weight of the vehicle or allow more power in the samesized design, and also allows a sensor to be added inside the battery cells for more effective monitoring. The first versions have been printed as cells containing patterned openings for thermal management. This uses a proprietary, multi-material, multi-layer approach in a parallel and dry process, instead of layer-on-layer printing or screen-printing, which is used to build solid-state battery cells. Those inherently wet processes require significant energy to remove unwanted solvents, and they are susceptible to poor printing quality and unreliable production. The first printed batteries with a lithium-metal anode have demonstrated successful cycling performance at a slower C/5 charging rate and are expected to achieve an energy density of 800-1000 Wh/L. Future directions Generative AI is being used to discover and produce new electrolytes for solidstate batteries, and to speed up the testing process. A tool called the Graph Networks for Materials Exploration (GNoME) was trained on data about the efficiency of 48,000 stable materials. This identified 2.2 million possible structures, many of them new. The AI then sifted through the list to find 321,000 potential materials that were then simulated at a molecular level to determine the functions to an accuracy of 11 megaelectron volts per atom (meV/atom). Of the stable structures, 736 materials have already been independently experimentally produced to verify the AI output. A hybrid approach couples experimental and simulation methods, enabling insights into the formation process and characteristics of battery interfaces at the molecular level, while using experimental data to improve the accuracy of the models. This uses AI to extract patterns from gigantic data sets for electrolyte design and battery life prediction. For testing, a tool called Coscientist, which is driven by GPT-4 generative AI, can autonomously design, plan and perform complex experiments with code execution and experimental automation. This is used to accelerate research across various tasks with autonomous experimental design and execution. 47 Uncrewed Systems Technology | April/May 2024 3D-printed battery cells (Image courtesy of Sakuu) The move to solid electrolytes creates the opportunity to use additive manufacturing to print battery cells… this can reduce the weight of the vehicle