Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog

Category: robotics/AI – Page 701

“Flying dragon” robot harnesses the “crazy hose” effect to fight fires

Japanese researchers have created and open-sourced a flying firefighting hose that levitates and steers itself to fight fires using its own water pressure as a two-part propulsion system, spraying water down onto fires and keeping operators safe.

The “flying dragon” system has two four-nozzle propulsion units built in – one at the end of the hose, one maybe 3 m (10 ft) back. Each of these can be thought of as something like a watery quadcopter – valves and swivels on each nozzle control flow and direction of thrust, allowing it to rise, balance and steer itself in the air the way a regular drone might … Well, two drones really, connected with a heavy rope and dragging a heavy tail.

A maximum flow rate of 6.6 liters (1.5 gal) per second gives pressure ratings up to 1 megapascal (145 psi). That’s enough to lift the hose some 2 m (6.6 ft) above the last thing it’s been draped on. The hose on the prototype at this point is just 4 m (13.2 ft) long, and runs back to a little control station trolley, where an operator stands and drives the thing.

Model scale versus domain knowledge in statistical forecasting of chaotic systems

Can machine learning predict chaos? This paper performs a large-scale comparison of modern forecasting methods on a giant dataset of 135 chaotic systems.

Chaos and unpredictability are traditionally synonymous, yet large-scale machine-learning methods recently have demonstrated a surprising ability to forecast chaotic systems well beyond typical predictability horizons. However, recent works disagree on whether specialized methods grounded in dynamical systems theory, such as reservoir computers or neural ordinary differential equations, outperform general-purpose large-scale learning methods such as transformers or recurrent neural networks. These prior studies perform comparisons on few individually chosen chaotic systems, thereby precluding robust quantification of how statistical modeling choices and dynamical invariants of different chaotic systems jointly determine empirical predictability.

AI Coscientist automates scientific discovery

A non-organic intelligent system has for the first time designed, planned and executed a chemistry experiment, Carnegie Mellon University researchers report in the journal Nature (“Autonomous chemical research with large language models”).

  • A non-organic intelligent system has successfully conducted a chemistry experiment, demonstrating a new approach to scientific research.
  • The system, named Coscientist, leverages large language models to streamline the experimental process, enhancing speed, accuracy, and efficiency.

    • Chinese brain warfare includes sleep weapons, thought control

    I dont know about sleep weapons, it s possible probably. More concerning to me, i read a paper 20+ years back about cell towers and cell phone frequencies as a possible tool for mind control, some way connected to frequency of human brain.

    China’s military is developing advanced psychological warfare and brain-influencing weapons as part of a new warfighting strategy, according to a report on People’s Liberation Army cognitive warfare.

    The report, “Warfare in the Cognitive Age: NeuroStrike and the PLA’s Advanced Psychological Weapons and Tactics,” was published earlier this month by The CCP Biothreats Initiative, a research group.

    “The PLA is at the forefront of incorporating advanced technologies such as artificial intelligence, brain-computer interfaces and novel biological weapons into its military strategies,” the think tank’s analysts concluded.

    A Comprehensive Study on Nanoparticle Drug Delivery to the Brain: Application of Machine Learning Techniques

    The delivery of drugs to specific target tissues and cells in the brain poses a significant challenge in brain therapeutics, primarily due to limited understanding of how nanoparticle (NP) properties influence drug biodistribution and off-target organ accumulation. This study addresses the limitations of previous research by using various predictive models based on collection of large data sets of 403 data points incorporating both numerical and categorical features. Machine learning techniques and comprehensive literature data analysis were used to develop models for predicting NP delivery to the brain. Furthermore, the physicochemical properties of loaded drugs and NPs were analyzed through a systematic analysis of pharmacodynamic parameters such as plasma area under the curve. The analysis employed various linear models, with a particular emphasis on linear mixed-effect models (LMEMs) that demonstrated exceptional accuracy. The model was validated via the preparation and administration of two distinct NP formulations via the intranasal and intravenous routes. Among the various modeling approaches, LMEMs exhibited superior performance in capturing underlying patterns. Factors such as the release rate and molecular weight had a negative impact on brain targeting. The model also suggests a slightly positive impact on brain targeting when the drug is a P-glycoprotein substrate.

    Researchers from Indiana University Unveil ‘Brainoware’: A Cutting-Edge Artificial Intelligence Technology Inspired by Brain Organoids and Silicon Chips

    The fusion of biological principles with technological innovation has resulted in significant advancements in artificial intelligence (AI) through the development of Brainoware. Developed by researchers at Indiana University, Bloomington, this innovative system leverages clusters of lab-raised brain cells to achieve elementary speech recognition and solve mathematical problems.

    The crux of this technological leap lies in the cultivation of specialized stem cells that mature into neurons—the fundamental units of the brain. While a typical human brain comprises a staggering 86 billion neurons interconnected extensively, the team managed to engineer a minute organoid, merely a nanometer wide. This tiny but powerful structure was connected to a circuit board through an array of electrodes, allowing machine-learning algorithms to decode responses from the brain tissue.

    Termed Brainoware, this amalgamation of biological neurons and computational circuits exhibited remarkable capabilities after a brief training period. It was discerned between eight subjects based on their diverse pronunciation of vowels with an accuracy rate of 78%. Impressively, Brainoware outperformed artificial networks in predicting the Henon map, a complex mathematical construct within chaotic dynamics.

    Mechanical intelligence simplifies control in terrestrial limbless locomotion

    To advance our overall understanding and discover principles of mechanical intelligence in limbless locomotion and to understand the potential role of bilateral actuation specifically in mechanical control, we took a comparative biological and robophysical approach using two complementary models: a biological model, the nematode C. elegans, and a robophysical model, a limbless robot incorporating a bilateral actuation scheme that permits programmable, dynamic, and quantifiable body compliance (Fig. 1B). This compliance governs the passive body-environment interactions in the horizontal plane that allow mechanical intelligence. Because separating neural and mechanical aspects of control is challenging in a freely locomoting living system, we used the robot as a model (22, 24, 49, 50) that then allowed mechanical intelligence to be isolated from active controls and to be systematically tuned and tested.

    Using comparisons between the kinematics and locomotor performance of our biological and robophysical models, we show that mechanical intelligence alone is sufficient for an open-loop limbless robot to reproduce locomotory behavior of nematodes. Mechanical intelligence simplifies controls in terrestrial limbless locomotion by taking advantage of passive body-environment interactions that enable heterogeneity negotiation, thereby stabilizing locomotion. Further, we show that a simple active behavior inspired by nematodes takes advantage of mechanical intelligence to enhance locomotion performance even further. Our method and results not only provide insight into the functional mechanism of mechanical intelligence in organismal limbless locomotion but also provide an alternative paradigm for limbless robot development that simplifies control in complex environments.