Explore the vast range of titles available.

First-person view (FPV) drone racing is a televised sport in which professional competitors pilot high-speed aircraft through a 3D circuit. Each pilot sees the environment from the perspective of their drone by means of video streamed from an onboard camera. Reaching the level of professional pilots with an autonomous drone is challenging because the robot needs to fly at its physical limits while estimating its speed and location in the circuit exclusively from onboard sensors 1 . Here we introduce Swift, an autonomous system that can race physical vehicles at the level of the human world champions. The system combines deep reinforcement learning (RL) in simulation with data collected in the physical world. Swift competed against three human champions, including the world champions of two international leagues, in real-world head-to-head races. Swift won several races against each of the human champions and demonstrated the fastest recorded race time. This work represents a milestone for mobile robotics and machine intelligence 2 , which may inspire the deployment of hybrid learning-based solutions in other physical systems. An autonomous system is described that combines deep reinforcement learning with onboard sensors collecting data from the physical world, enabling it to fly faster than human world champion drone pilots around a race track.

Aircraft, and the aviation ecosystem in which they operate, are shaped by complex trades among technical requirements, economics and environmental concerns, all built on a foundation of safety. This Perspective explores the requirements of battery-powered aircraft and the chemistries that hold promise to enable them. The difference between flight and terrestrial needs and chemistries are highlighted. Safe, usable specific energy rather than cost is the major constraint for aviation. We conclude that battery packs suitable for flight with specific energy approaching 600 watt hours per kilogram may be achievable in the next decade given sufficient investment targeted at aeronautical applications. The economic, technical, environmental and safety requirements of battery-powered aircraft are considered, and promising technologies and future prospects for battery innovation are discussed.

Robot designs can take many inspirations from nature, where there are many examples of highly resilient and fault-tolerant locomotion strategies to navigate complex terrains by recruiting multi-functional appendages. For example, birds such as Chukars and Hoatzins can repurpose wings for quadrupedal walking and wing-assisted incline running. These animals showcase impressive dexterity in employing the same appendages in different ways and generating multiple modes of locomotion, resulting in highly plastic locomotion traits which enable them to interact and navigate various environments and expand their habitat range. The robotic biomimicry of animals’ appendage repurposing can yield mobile robots with unparalleled capabilities. Taking inspiration from animals, we have designed a robot capable of negotiating unstructured, multi-substrate environments, including land and air, by employing its components in different ways as wheels, thrusters, and legs. This robot is called the Multi-Modal Mobility Morphobot, or M4 in short. M4 can employ its multi-functional components composed of several actuator types to (1) fly, (2) roll, (3) crawl, (4) crouch, (5) balance, (6) tumble, (7) scout, and (8) loco-manipulate. M4 can traverse steep slopes of up to 45 deg. and rough terrains with large obstacles when in balancing mode. M4 possesses onboard computers and sensors and can autonomously employ its modes to negotiate an unstructured environment. We present the design of M4 and several experiments showcasing its multi-modal capabilities. The biomimicry of animals’ appendage repurposing can be applied to robot designs, resulting in unparalleled capabilities. Sihite et al. report a Multi-Modal Mobility Morphobot (M4) that negotiate unstructured, multi-substrate environments, including land and air, by employing its components in different ways as wheels, thrusters, and legs.

Accurate flight trajectory prediction is a crucial and challenging task in air traffic control, especially for maneuver operations. Modern data-driven methods are typically formulated as a time series forecasting task and fail to retain high accuracy. Meantime, as the primary modeling method for time series forecasting, frequency-domain analysis is underutilized in the flight trajectory prediction task. In this work, an innovative wavelet transform-based framework is proposed to perform time-frequency analysis of flight patterns to support trajectory forecasting. An encoder-decoder neural architecture is developed to estimate wavelet components, focusing on the effective modeling of global flight trends and local motion details. A real-world dataset is constructed to validate the proposed approach, and the experimental results demonstrate that the proposed framework exhibits higher accuracy than other comparative baselines, obtaining improved prediction performance in terms of four measurements, especially in the climb and descent phase with maneuver control. Most importantly, the time-frequency analysis is confirmed to be effective to achieve the flight trajectory prediction task. Accurate flight trajectory prediction can be a challenging task in air traffic control, especially for maneuver operations. Here, authors develop a time-frequency analysis based on an encoder-decoder neural architecture to estimate wavelet components and model global flight trends and local motion details.

Carbon fiber-reinforced polymer (CFRP) easily realizes the integrated manufacturing of components with high specific strength and stiffness, and it has become the preferred material in the aerospace field. Grinding is the key approach to realize precision parts and matching the positioning surface for assembly and precision. Hygroscopicity limits the application of flood lubrication in CFRP grinding, and dry grinding leads to large force, surface deterioration, and wheel clogging. To solve the above technical bottleneck, this study explored the grindability and frictional behavior of CNT biological lubricant MQL through grinding experiments and friction-wear tests. Results showed that the CNT biological lubricant reduced the friction coefficient by 53.47% compared with dry condition, showing optimal and durable antifriction characteristics. The new lubrication was beneficial to suppressing the removal of multifiber block debris, tensile fracture, and tensile-shear fracture, with the advantages of tribological properties and material removal behavior, the tangential and normal grinding force, and the specific grinding energy were reduced by 40.41%, 31.46%, and 55.78%, respectively, compared with dry grinding. The proposed method reduced surface roughness and obtained the optimal surface morphology by preventing burrs, fiber pull-out, and resin smearing, and wheel clogging was prevented by temperature reduction and lubricating oil film formation. S a and S q of the CNT biological lubricant were reduced by 8.4% and 7.9%, respectively, compared with dry grinding. This study provides a practical basis for further application of CNT biological lubricant in CFRP grinding.

Efficiently navigating a superpressure balloon in the stratosphere 1 requires the integration of a multitude of cues, such as wind speed and solar elevation, and the process is complicated by forecast errors and sparse wind measurements. Coupled with the need to make decisions in real time, these factors rule out the use of conventional control techniques 2 , 3 . Here we describe the use of reinforcement learning 4 , 5 to create a high-performing flight controller. Our algorithm uses data augmentation 6 , 7 and a self-correcting design to overcome the key technical challenge of reinforcement learning from imperfect data, which has proved to be a major obstacle to its application to physical systems 8 . We deployed our controller to station Loon superpressure balloons at multiple locations across the globe, including a 39-day controlled experiment over the Pacific Ocean. Analyses show that the controller outperforms Loon’s previous algorithm and is robust to the natural diversity in stratospheric winds. These results demonstrate that reinforcement learning is an effective solution to real-world autonomous control problems in which neither conventional methods nor human intervention suffice, offering clues about what may be needed to create artificially intelligent agents that continuously interact with real, dynamic environments. Data augmentation and a self-correcting design are used to develop a reinforcement-learning algorithm for the autonomous navigation of Loon superpressure balloons in challenging stratospheric weather conditions.

Humankind's aerospace aspirations are placing unprecedented demands on vehicle propulsion systems. Advanced structural ceramics are playing a key role in addressing these challenges.

Additive manufacturing methods 1 – 4 using static and mobile robots are being developed for both on-site construction 5 – 8 and off-site prefabrication 9 , 10 . Here we introduce a method of additive manufacturing, referred to as aerial additive manufacturing (Aerial-AM), that utilizes a team of aerial robots inspired by natural builders 11 such as wasps who use collective building methods 12 , 13 . We present a scalable multi-robot three-dimensional (3D) printing and path-planning framework that enables robot tasks and population size to be adapted to variations in print geometry throughout a building mission. The multi-robot manufacturing framework allows for autonomous three-dimensional printing under human supervision, real-time assessment of printed geometry and robot behavioural adaptation. To validate autonomous Aerial-AM based on the framework, we develop BuilDrones for depositing materials during flight and ScanDrones for measuring the print quality, and integrate a generic real-time model-predictive-control scheme with the Aerial-AM robots. In addition, we integrate a dynamically self-aligning delta manipulator with the BuilDrone to further improve the manufacturing accuracy to five millimetres for printing geometry with precise trajectory requirements, and develop four cementitious–polymeric composite mixtures suitable for continuous material deposition. We demonstrate proof-of-concept prints including a cylinder 2.05 metres high consisting of 72 layers of a rapid-curing insulation foam material and a cylinder 0.18 metres high consisting of 28 layers of structural pseudoplastic cementitious material, a light-trail virtual print of a dome-like geometry, and multi-robot simulations. Aerial-AM allows manufacturing in-flight and offers future possibilities for building in unbounded, at-height or hard-to-access locations. An additive manufacturing method using a team of autonomous aerial robots allows for scalable and adaptable three-dimensional printing, and is used to deposit building materials during flight.

Bulk equiatomic (Hf-Ta-Zr-Ti)C and (Hf-Ta-Zr-Nb)C high entropy Ultra-High Temperature Ceramic (UHTC) carbide compositions were fabricated by ball milling and Spark Plasma Sintering (SPS). It was found that the lattice parameter mismatch of the component monocarbides is a key factor for predicting single phase solid solution formation. The processing route was further optimised for the (Hf-Ta-Zr-Nb)C composition to produce a high purity, single phase, homogeneous, bulk high entropy material (99% density); revealing a vast new compositional space for the exploration of new UHTCs. One sample was observed to chemically decompose; indicating the presence of a miscibility gap. While this suggests the system is not thermodynamically stable to room temperature, it does reveal further potential for the development of new in situ formed UHTC nanocomposites. The optimised material was subjected to nanoindentation testing and directly compared to the constituent mono/binary carbides, revealing a significantly enhanced hardness (36.1 ± 1.6 GPa,) compared to the hardest monocarbide (HfC, 31.5 ± 1.3 GPa) and the binary (Hf-Ta)C (32.9 ± 1.8 GPa).

Finite-element models are used to identify a material geometry that achieves the theoretical bounds on isotropic elastic stiffness—a combination closed-cell cubic and octet foam. Know your limits Lattice-like arrangements of trusses are a well-known way of achieving lightweight structures with high stiffness and strength. They are just one example of a broader family of materials—sometimes termed mechanical metamaterials—in which structural geometry is harnessed to achieve enhanced combinations of properties. Yet theory predicts limits on these properties that have yet to be attained. Jonathan Berger and colleagues use a combination of theoretical and optimization techniques to identify a material geometry that reaches the theoretical limits for isotropic elastic stiffness. Such a structure should be achievable, for example, by harnessing recent advances in 3D printing. A wide variety of high-performance applications 1 require materials for which shape control is maintained under substantial stress, and that have minimal density. Bio-inspired hexagonal and square honeycomb structures and lattice materials based on repeating unit cells composed of webs or trusses 2 , when made from materials of high elastic stiffness and low density 3 , represent some of the lightest, stiffest and strongest materials available today 4 . Recent advances in 3D printing and automated assembly have enabled such complicated material geometries to be fabricated at low (and declining) cost. These mechanical metamaterials have properties that are a function of their mesoscale geometry as well as their constituents 3 , 5 , 6 , 7 , 8 , 9 , 10 , 11 , 12 , leading to combinations of properties that are unobtainable in solid materials; however, a material geometry that achieves the theoretical upper bounds for isotropic elasticity and strain energy storage (the Hashin–Shtrikman upper bounds) has yet to be identified. Here we evaluate the manner in which strain energy distributes under load in a representative selection of material geometries, to identify the morphological features associated with high elastic performance. Using finite-element models, supported by analytical methods, and a heuristic optimization scheme, we identify a material geometry that achieves the Hashin–Shtrikman upper bounds on isotropic elastic stiffness. Previous work has focused on truss networks and anisotropic honeycombs, neither of which can achieve this theoretical limit 13 . We find that stiff but well distributed networks of plates are required to transfer loads efficiently between neighbouring members. The resulting low-density mechanical metamaterials have many advantageous properties: their mesoscale geometry can facilitate large crushing strains with high energy absorption 2 , 14 , 15 , optical bandgaps 16 , 17 , 18 , 19 and mechanically tunable acoustic bandgaps 20 , high thermal insulation 21 , buoyancy, and fluid storage and transport. Our relatively simple design can be manufactured using origami-like sheet folding 22 and bonding methods.