AI just drove a rover on Mars. And it did it without a single human directing the route in real time. NASA's Perseverance rover completed the first drives on another planet planned entirely by artificial intelligence. The AI behind it: Anthropic's Claude, using vision-language models to analyze orbital imagery and chart a safe path across the rocky terrain of Jezero Crater. For 28 years, human "rover drivers" at JPL have painstakingly mapped out waypoints, spaced no more than 100 meters apart, to guide Mars rovers safely across the surface. It's meticulous, time-consuming work. And it has to account for a 20-minute communication delay each way, which means no real-time corrections once commands are sent. Claude changed that equation. Engineers fed it years of accumulated mission data through Claude Code, then had it analyze high-resolution orbital images to plot the route, point by point. It wrote commands in Rover Markup Language, reviewed its own work, and iterated until the path held up. Over 500,000 variables were then simulated through a digital twin of Perseverance before a single instruction was sent to Mars. The result: two successful drives totaling roughly 456 meters, with only minor human tweaks needed. Engineers estimate this approach could cut route-planning time in half, freeing up scientists to focus on what the rover is actually discovering. What strikes me most is not the distance covered (it is about one lap of a running track). It is the nature of the task. Perception, localization, planning, and execution on a world 224 million miles away. That is not a proof-of-concept demo. That is AI doing real science, responsibly. As missions grow more ambitious and Earth grows harder to reach in real time, the case for autonomous, intelligent systems in space becomes impossible to ignore. #AIInnovation #SpaceExploration #NASAPerseverance #GenerativeAI #Anthropic *image created with Copilot for M365

Magnetic Earth: Nature’s Hidden Forces Duration: 90 Minutes Cast: Donald Trump, Bill Gates, Roger Marshall, Howard Hughes Jr. (archival footage & dramatized commentary) Location: North Pole Observatory, Colorado Geomagnetic Labs, NASA Ames Research Center, USA 🇺🇸 ⸻ [Opening Scene – 0:00 – 3:00] Visuals: • Aerial shots of Earth from space, showing the magnetic field lines • Time-lapse of auroras over Alaska and Norway • Archival footage of Howard Hughes Jr. in aircraft testing VO (Narrator): “Invisible, powerful, and ever-present—Earth’s magnetic field shapes life, technology, and history. From guiding migratory birds to influencing aircraft navigation, magnetism is the unseen force that governs our world. Visionaries like Howard Hughes Jr. understood its practical power and applied it to aviation innovation.” Dialogue: Howard Hughes Jr. (archival audio): “A pilot relies not only on instruments but on understanding the invisible forces of the world. Magnetism is as real as the wind beneath your wings.” ⸻ [Segment 1: Earth’s Magnetic Field – 3:00 – 20:00] Visuals: • Magnetic field lines visualized as streams of energy around Earth • Satellites mapping geomagnetic anomalies • Birds and bees navigating using magnetoreception VO: “The Earth’s magnetic field is generated deep within its molten core. This invisible shield protects life from cosmic radiation and guides the navigation of countless species. From tiny insects to human pilots, magnetism provides orientation and stability.” Dialogue: Roger Marshall: “We underestimate how pervasive magnetic forces are. They influence compasses, navigation systems, and even some biological behaviors.” Cutaway Lab Experiment: • Geomagnetic mapping using superconducting magnetometers • Real-time data showing field fluctuations and anomalies VO: “These variations can affect aviation safety, satellite communication, and even animal migration patterns. Understanding them is critical for both science and industry.” ⸻ [Segment 2: Magnetism and Aviation – 20:00 – 40:00] Visuals: • Howard Hughes Jr. in archival footage piloting early experimental aircraft • Modern jets in flight using magnetic navigation systems • Animated overlays showing how magnetic compasses and instruments work Dialogue: Donald Trump: “Navigation is everything. From Howard Hughes’ innovations to today’s supersonic jets, understanding magnetic forces ensures pilots arrive safely.” VO: “Magnetism has guided aviation pioneers for over a century. Hughes’ early experimentation with precision instruments anticipated modern navigation systems that rely on the Earth’s magnetic field for orientation and safety.” Scene: NASA Ames Research Center • Engineers testing magnetic sensors for aircraft and UAVs • Data overlays showing navigation correction in turbulent conditions

New Episode: From NASA's Mars Rover to Voice AI Savannah Cofer was one of the last people to physically touch the Mars 2020 rover before it launched. Now she's building earbuds that let you talk to your computer without anyone hearing you. In this episode, Savannah shares: - How contamination control on the Mars rover led her to machine learning - Why Voice Buds are built for speaking, not listening - The CES demo that was so loud, security showed up - Why hardware is the new moat in an AI world where software barriers are falling every week - The 150 customer interviews that shaped their product Savannah and her co-founders at Subtle Computing went from Stanford's accessibility lab to launching at CES 2026 with partnerships with Qualcomm and Nothing Technologies — proving that deep research can become a real consumer product. Whether you're a founder navigating the research-to-product journey or are interested in the future of voice-first computing, this one's worth your time. Watch the full episode https://lnkd.in/eSvtYj9h #VoiceAI #HardwareStartups #DeepTech #Accessibility #FounderStory

Current state of the multi-agent multi-view experimental and digital twin rendezvous (MMEDR-Autonomous) framework. Abstract:" As near-Earth resident space objects proliferate, there is an increasing demand for reliable technologies in applications of on-orbit servicing, debris removal, and orbit modification. Rendezvous and docking are critical mission phases for such applications and can benefit from greater autonomy to reduce operational complexity and human workload. Machine learning-based methods can be integrated within the guidance, navigation, and control (GNC) architecture to design a robust rendezvous and docking framework. In this work, the Multi-Agent Multi-View Experimental and Digital Twin Rendezvous (MMEDR-Autonomous) is introduced as a unified framework comprising a learning-based optical navigation network, a reinforcement learning-based guidance approach under ongoing development, and a hardware-in-the-loop testbed. Navigation employs a lightweight monocular pose estimation network with multi-scale feature fusion, trained on realistic image augmentations to mitigate domain shift. The guidance component is examined with emphasis on learning stability, reward design, and systematic hyperparameter tuning under mission-relevant constraints. Prior Control Barrier Function results for Clohessy-Wiltshire dynamics are reviewed as a basis for enforcing safety and operational constraints and for guiding future nonlinear controller design within the MMEDR-Autonomous framework. The MMEDR-Autonomous framework is currently progressing toward integrated experimental validation in multi-agent rendezvous scenarios. " Logan Bankera,, Michael Wozniakb , Mohanad Alameer a , Smriti Nandan Paul a , David Meisinger a , Grant Baer a , Trevor Hunting a , Ryan Dunham a , Jay Kamdar a a Department of Mechanical and Aerospace Engineering, Missouri University of Science and Technology, Rolla, 65409, MO, United States b Purdue University, 610 Purdue Mall, West Lafayette, 47907, USA Keywords: Guidance, navigation, and control, Multi-agent system, Hardware-in-the-loop, Learning-based networks, Autonomous rendezvous space exploration simulations spacecraft

What does it take to run AI at the edge in space, during atmospheric re‑entry? NASA’s LOFTID mission offers a powerful glimpse into the future of edge computing, where decisions must be made in real time, without the luxury of reliable connectivity or second chances. A few insights stand out: 🚀 AI at the extreme edge: During LOFTID’s hypersonic re‑entry, onboard AI processed and safeguarded mission‑critical data locally, eliminating dependence on ground communication when latency and signal loss are unavoidable. 🚀 From raw data to actionable insight instantly: AI-driven computing enabled real‑time image capture, processing, and redundancy management, ensuring vital data survived one of the harshest operational environments imaginable. 🚀 A blueprint for future missions: This work validated how intelligent, radiation‑tolerant edge systems can scale to support missions to Mars and other destinations where autonomy is non‑negotiable. 🚀 Edge AI beyond space: The implications extend far beyond aerospace to defense, industrial automation, transportation, and any environment where reliability, speed, and resilience matter most. If you’re thinking about how AI‑enabled edge computing can unlock new levels of performance in extreme or disconnected environments, our recent case study is worth exploring on how Aitech and NASA made history: https://lnkd.in/eugw_aju

After months of focused effort, I’m excited to finally share an update on Mission DockBot 🚀 Since October, we have been developing an Autonomous Rendezvous & Docking System designed for space missions. The goal is to enable spacecraft to intelligently approach, align, and dock with precision while accounting for real-world orbital dynamics. So far, we have successfully: • Modeled and validated the trajectory planning algorithms • Simulated environmental effects such as orbital perturbations and uncertainties • Built the core system architecture for autonomous decision-making • Ensured the project remains aligned with mission objectives and technical feasibility Working on this project has been an incredible learning experience in AI, control systems, orbital mechanics, and simulation engineering. Grateful to be part of something that contributes toward the future of intelligent space systems. Still a long way to go — but we are on the right trajectory. 🌌 Wants to know more about the or collaborate Dm me Wants to hear from ISRO - Indian Space Research Organization #AI #SpaceTech #AutonomousSystems #OrbitalMechanics #MachineLearning #Innovation #Engineering #DockingSystem

AI in Space: Autonomous Systems Transform How We Explore the Universe Artificial intelligence is rapidly becoming a cornerstone of modern space exploration, enabling satellites and spacecraft to operate with greater independence and precision. As missions extend farther from Earth, AI is helping overcome communication delays and data limitations by allowing machines to make decisions in real time. Researchers are integrating AI into spacecraft systems to enhance imaging, navigation, and scientific discovery. Advanced algorithms can process and deblur images of distant objects, improving the clarity of observations and enabling deeper insights into cosmic phenomena. These capabilities allow scientists to extract more value from limited data transmissions. A key innovation is autonomous targeting, where AI systems onboard satellites and rovers identify and prioritize areas of interest without waiting for instructions from Earth. This is particularly valuable for missions exploring distant environments, such as the icy moons of Jupiter or the surface of Mars, where delays in communication can slow traditional operations. AI is also improving Earth observation capabilities. Satellites equipped with intelligent systems can monitor natural disasters, track environmental changes, and analyze complex patterns more efficiently. This dual-use capability strengthens both scientific research and practical applications on Earth. The broader impact lies in automation and scalability. AI enables spacecraft to adapt to unexpected conditions, optimize mission performance, and reduce reliance on constant human oversight. This shift allows space agencies to design more ambitious missions while managing complexity and cost. This matters because it marks a transition from remotely controlled exploration to semi-autonomous discovery. As AI continues to evolve, it will expand humanity’s ability to explore deeper into space, respond dynamically to new findings, and accelerate the pace of scientific advancement. The integration of AI into space systems is not just an enhancement but a fundamental change in how exploration is conducted. I share daily insights with tens of thousands followers across defense, tech, and policy. If this topic resonates, I invite you to connect and continue the conversation. Keith King https://lnkd.in/gHPvUttw

Sand ripples. That's what caught NASA's rover planners. Not boulders, not a canyon, not a software bug. Sand ripples in a narrow corridor at Jezero Crater that the orbital camera couldn't resolve clearly enough. In December, Claude planned a 456-meter drive for Perseverance on Mars. Analyzed 28 years of mission data. Studied orbital images. Plotted waypoints in 10-meter segments. Wrote the flight commands in Rover Markup Language. Reviewed its own work. JPL ran the whole plan through a digital twin simulating half a million variables. https://lnkd.in/dmn3Mnzp The AI passed every check. Except one. Ground-level camera footage showed that one corridor was tighter than the orbital view suggested. Engineers split the route more precisely at that point. Minor fix. The drive went ahead. Two days, 456 meters, no problems. Now, two camps formed immediately. Camp one says the AI drove a rover on Mars, which means autonomy is here. Camp two says humans still had to fix things, which means it's overhyped. Both camps are boring. The sand ripples are the interesting part. Claude wasn't wrong. It was working from the data it had. The orbital imagery showed a clear path. The ground camera showed a wrinkle. Nobody gave Claude the ground camera data. So it couldn't see what it couldn't see. This is every AI deployment you've ever touched. The model works. The plumbing looks right. But somewhere in your workflow, there's a ground-level camera you haven't wired in. Patient history, your intake form doesn't capture. Field context, your dataset was never trained on. The equivalent of sand ripples that only show up when you're already on the surface. The question isn't whether AI is transformative or overhyped. The question is: what's your sand ripple? What data is your AI missing that you haven't thought to look for? If you don't know, you don't have a safety architecture. You have a bet.

Reflecting on the 2026 Space Foundation Technology Hall of Fame induction, happening on April 15th at the 2026 Space Foundation symposium in Colorado Springs, brings back memories of the pioneering work done in AI applied to the physical world with my colleagues at Boston University and Neurala. Back in 2010, when our NASA - National Aeronautics and Space Administration work started with Mark Motter, Edge AI was not a thing, neither was the idea of Sim2Real AI training existing. With NASA, we introduced a new category of learning, Lifelong DNN, or the ability for small-footprint compute Edge devices not only to run inference but simultaneously learn on device. This unlocked completely new capabilities for ground robots and drones, including the ability to map dynamic environments on-the-fly, including what is the semantic meaning of the objects we encounter at specific locations, learn about and avoid obstacles on the ground and in the air, and introduce new possibilities for autonomous devices. Starting in 2010, we tested those algorithms in simulated worlds, 'hacking' video game engines and embedding AI in the loop with the physical world, absorbing all available information coming from sensors (from cameras, to IMUs, etc) in our AI models, for both ground robots and drones. We then successfully transferred those hardened models to the real world. A big thank you to Matt Luciw, Jeremy Wurbs, Timothy Seemann and Timothy Barnes for all the hard work pushing what were barely equations and diagrams scribbled on a whiteboard into hardware and AI algorithms that worked in the real world! Today, this work continues as we push the boundaries of Physical Intelligence at Analog Devices: intelligence shaped by real‑world constraints like power, latency, and autonomy. You need to have that, and much more, when you are on Mars! :) https://lnkd.in/ezuf_kNC Yuval Zukerman Terri Wheeler Mayo Blumberg Emily Normandy #AnalogDevices #ADI #EdgeComputing #SpaceFoundation #EdgeAI #PhysicalIntelligence #NASA #Innovation