Lifeboat Foundation

The past decade has witnessed the great success of deep neural networks in various domains. However, deep neural networks are very resource-intensive in terms of energy consumption, data requirements, and high computational costs. With the recent increasing need for the autonomy of machines in the real world, e.g., self-driving vehicles, drones, and collaborative robots, exploitation of deep neural networks in those applications has been actively investigated. In those applications, energy and computational efficiencies are especially important because of the need for real-time responses and the limited energy supply. A promising solution to these previously infeasible applications has recently been given by biologically plausible spiking neural networks. Spiking neural networks aim to bridge the gap between neuroscience and machine learning, using biologically realistic models of neurons to carry out the computation. Due to their functional similarity to the biological neural network, spiking neural networks can embrace the sparsity found in biology and are highly compatible with temporal code. Our contributions in this work are: (i) we give a comprehensive review of theories of biological neurons; (ii) we present various existing spike-based neuron models, which have been studied in neuroscience; (iii) we detail synapse models; (iv) we provide a review of artificial neural networks; (v) we provide detailed guidance on how to train spike-based neuron models; (vi) we revise available spike-based neuron frameworks that have been developed to support implementing spiking neural networks; (vii) finally, we cover existing spiking neural network applications in computer vision and robotics domains. The paper concludes with discussions of future perspectives.

Keywords: spiking neural networks, biological neural network, autonomous robot, robotics, computer vision, neuromorphic hardware, toolkits, survey, review.

Go to:

MIT’s soft drone flies and grasps, swiftly picking up a bottle in a demo video:

Interestingly, the drone’s new capabilities allow it to catch objects that are moving at speeds of up to 0.3 meters per second.

Continue reading “MIT’s new drone with flying claws grabs objects mid-air at high speed” »

Three days ago, an old Falcon 9 booster crashed and exploded while landing on a drone ship during its 23rd landing.

The FAA has cleared SpaceX to launch the Falcon 9 while the investigation into Wednesday’s mishap wraps up.

Thank you @LiamVickersAnimation made us such a wonderful animated series, we wish you good luck in your life in the meantime, enjoy the movie of all the epi…

The Korea Institute of Energy Research (KIER) has developed a redox-active metal-organic hybrid electrode material (SKIER-5) for Li batteries that remains stable in cold conditions as low as minus 20 degrees Celsius. By addressing the limitations of graphite as an anode material of conventional Li batteries under freezing conditions, SKIER-5 has the potential to be a superior alternative. This novel material can be used in Li batteries for a variety of applications, including electric vehicles, drones, and ultra-small electronic devices, even at low temperatures.

Currently, graphite is the conventional material used for anodes in lithium-ion batteries due to its thermodynamic stability and low cost. However, batteries with graphite anodes have significant drawbacks: their storage capacity sharply decreases at subzero temperatures, and dendrites can form on the anode surface during charging. This can lead to thermal runaway and potential explosions.

A research team led by Dr. Jungjoon Yoo, Dr. Kanghoon Yim, and Dr. Hyunuk Kim at KIER has developed a redox-active conductive metal-organic framework called “SKIER-5.” This framework is assembled from a trianthrene-based organic ligand and nickel ions. SKIER-5 exhibited a discharge capacity five times higher than that of graphite in subzero environments.

A team of researchers at Delft University of Technology has developed a drone that flies autonomously using neuromorphic image processing and control based on the workings of animal brains. Animal brains use less data and energy compared to current deep neural networks running on GPUs (graphic chips). Neuromorphic processors are therefore very suitable for small drones because they don’t need heavy and large hardware and batteries.

The results are extraordinary: during flight the drone’s deep neural network processes data up to 64 times faster and consumes three times less energy than when running on a GPU. Further developments of this technology may enable the leap for drones to become as small, agile, and smart as flying insects or birds.

Photo of the “neuromorphic drone” flying over a flower pattern. It illustrates the visual inputs the drone receives from the neuromorphic camera in the corners. Red indicates pixels getting darker, green indicates pixels getting brighter. (Image: TU Delft)

SpaceX is working towards the goal of landing both the super heavy booster and Starship on a drone ship in the ocean, which has the potential to revolutionize space travel and support their mission for greater sustainability and reusability Questions to inspire discussion What is SpaceX’s goal for landing the super heavy booster and Starship?

Unmanned aerial vehicles (UAVs), commonly known as drones, are now used to capture images and carry out a wide range of missions in outdoor environments. While there are now several UAV designs with different advantages and characteristics, most conventional aerial robots are underactuated, meaning that they have fewer independent actuators than their degrees of freedom (DoF).

Underactuated systems are often more cost-effective and can be controlled using simpler control strategies than overactuated systems (i.e., robots that have more independent actuators than their DoF). Nonetheless, they are often less reliable and not as capable of precisely controlling their position and orientation.

Researchers at Tecnalia’s Basque Research and Technology Alliance (BRTA) in Spain recently developed a new overactuated aerial robot that can independently control the position and orientation of its main body. This robot, introduced in a paper published in Robotics and Autonomous Systems, has four quadrotors that cooperatively carry its central body.

As the Pentagon makes a push toward scaling production of autonomous systems and weapons, Anduril Industries is accelerating its own manufacturing capabilities through a new software-based production hub called Arsenal.

The California-based defense technology company announced Wednesday it will build the first Arsenal facility in the U.S., using funding from a recent $1.5 billion Series F investment round. Chris Brose, Anduril’s chief strategy officer, told reporters the firm’s goal is to consolidate manufacturing in order to “hyperscale” production across its product lines, including uncrewed combat drones and autonomous underwater vehicles.

“When we say hyperscale, we mean the ability to produce tens of thousands of a given system,” he said in a briefing. “This is the target that we’re setting for ourselves right now.”