Hardware Podcasts

198. Scared of Turbulence? AI ~Might~ Fix That

The Next Byte Podcast · 29 Oct 2024 · 23m

What this episode covers

Hosts Daniel and Forbode discuss a new algorithm called Falcon, developed by researchers from Caltech and Nvidia, designed to help aircraft and drones sense and counteract turbulence. The team uses Fourier methods to model wind energy shifts as waves, allowing the AI to stabilize a wing in a wind tunnel within nine minutes of training. The hosts compare this adaptive control system to emerging automotive technologies that adjust ride stiffness for road bumps, noting the potential for future commercial aviation and UAV applications.

Most relevant to

early-stage companies

Topics

Mentioned

Companies
Caltech, Nvidia, Singapore Airlines, Amazon
Products
Falcon

Related episodes from other shows

Other shows on the same ground, closest first.

  • EE Times Current · 15m

    Nvidia’s GTC | Boeing 737 Max | AI Special Projects

    Junko Yoshida, global coeditor of Aspencore Media, discusses the Boeing 737 Max crashes with editor-in-chief Brian Santo, attributing the failures to Boeing's decision to retrofit engines on an existing airframe and the resulting addition of the MCAS software as an afterthought. Rick Merritt reports on Nvidia's GTC conference, noting the absence of a 7nm GPU announcement, and previews an AI special project featuring insights from Dean Caiman of Cayman AI and HPE regarding the challenges of sensor fusion and neural network models. The episode also covers the U.S. Exascale supercomputer contract awarded to Intel and Cray, and a patent infringement verdict in the Qualcomm versus Apple case.

    Embedded firmwareAI & machine learning
    6 chapters

  • Hackaday Podcast · 53m

    Ep 122: Faster Than Wind Travel, Sisyphish, ALU Desktop Calculator, and Mice in Space

    Hackaday editors Mike Szczys and Elliot Williams discuss a variety of hardware hacks, including Rick Cavallero's Blackbird wind-powered car that travels faster than the wind, Zach Frew's 'Sisyphish' sand-drawing aquarium plotter, and Simon Boke's desktop calculator built around a 74LS181 ALU chip. They also cover Jonathan Gleik's use of a camera gimbal to stabilize his soldering iron against hand tremors, Deborah Ansel's modular triangle PCB art, and Robin Debrulli's low-temperature desoldering technique using bismuth and tin. The episode concludes with commentary on Ford's vehicle-to-home power capabilities and NASA's artificial gravity research using mice on the ISS.

    Mechanical engineeringAnalog & power electronics
    7 chapters

  • Hard Tech Podcast · 41m

    Colby Harvey: Drones for Aircraft Inspections, Boeing, Air Force, Air Safety | Hard Tech Podcast #13

    Colby Harvey, CEO of UltraView, explains how his startup uses autonomous drones equipped with 3D LiDAR and AI to inspect aircraft for defects like stress fractures and lightning strikes. He discusses the company's design process with Hatch Duo, its customers including Boeing and the U.S. Air Force, and the regulatory challenges of FAA approval for outdoor inspections. Harvey also addresses the impact of automation on maintenance jobs, the technical failures behind the Boeing 737 Max crashes, and the future of hardware versus software in the tech industry.

    Robotics & autonomyAI & machine learning
    8 chapters

  • Embedded.fm · 50m

    234: The Good Word About AI

    Dusty Franklin, a developer evangelist at NVIDIA, discusses the Jetson TX2, a low-power embedded module designed to bring AI and computer vision to edge devices like robots and drones. He explains how the platform's GPU enables parallel processing for tasks such as autonomous navigation, object detection, and reinforcement learning, often using simulation environments like ROS Gazebo to train agents safely. Franklin also details the 'Two Days to a Demo' tutorial, which guides developers through deploying and retraining neural networks using NVIDIA's TensorRT and DIGITS tools, and mentions the Redtail Project as an example of autonomous drone navigation.

    Embedded firmwareMicrocontrollers & dev platforms
    8 chapters

  • Hackaday Podcast · 48m

    Ep088: Flywheel Trebuchet, Thieving Magpies, Hero Engines, and Hypermiling

    Hackaday editors Mike Szczys and Elliot Williams discuss a variety of hardware hacks, including Tom Stanton’s flywheel trebuchet that throws tennis balls at 180 mph, Hans Forsberg’s system for training magpies to collect bottle caps, and Sarah Adkins’ Go board sequencer that uses OpenCV to generate music. They also cover Lou and Day’s hypermiling article, which details aerodynamic and weight-reduction techniques for improving fuel economy, and the announcement of the Sun's 25th solar cycle.

    Mechanical engineeringAI & machine learning
    6 chapters

The summary, chapters and topics are generated from the episode’s audio by this directory, not written by The Next Byte Podcast. The episode itself belongs to its publisher.