Kalin MLADENOV

France created a solid-state rocket engine that works without combustion — changing how we launch satellites forever In a quiet aerospace lab outside Toulouse, French engineers have developed something that may transform spaceflight from the ground up — a solid-state plasma propulsion engine that accelerates spacecraft without combustion, without moving parts, and without conventional fuel. It's not just a new engine — it's a new category of propulsion. This innovation is built on an ionized gas loop called a rotating detonation plasma disk, which uses magnetic fields to confine and spin superheated ions. Unlike chemical rockets that burn propellant in a loud, violent flame, this system moves particles using electric fields, producing quiet but continuous thrust with almost no mechanical wear. The core advantage? Precision. Because it’s electromagnetic, it can throttle, steer, or shut off instantly — crucial for satellite positioning, station-keeping, and space debris avoidance. In tests, it delivered stable thrust for over 1,000 hours with no degradation, far outpacing traditional ion thrusters. Even more impressive: it works in near vacuum, at low temperatures, and needs no ignition — meaning satellites can use it for years without refueling. The French team designed it to run on xenon, but it’s also being adapted for argon or krypton — making it cheaper and more versatile than current systems. This could drastically lower the cost of operating low-Earth orbit constellations, deep-space science probes, and even Mars-bound cargo ships. Unlike rocket launches, which are short and explosive, this tech allows long, efficient burns over months — ideal for modern space infrastructure. France’s space agency is already partnering with EU firms to integrate this engine into next-gen micro-launchers and orbital service vehicles — making combustion-free satellite propulsion a reality.

317

40 commentaires

What brings MetOp-SG A to life? Let's open the satellite 👀 MetOp-SG A1🛰️ is now inside the Ariane 6 rocket and is almost ready for the launch 🚀. This marvel of engineering is packed with an incredible array of instruments, each designed to give us insights into our planet's atmosphere and climate. Airbus and its subsidiaries across Europe have contributed to some of these sophisticated instruments by providing key equipment. Discover how ARIANESPACE European Space Agency - ESA EUMETSAT #VA264

287

5 commentaires

[#SagaSafran🕰️] CFM International (CFM) unveils the #CFMRISE program. This innovative technology development program for sustainable engines is a key lever to help the aviation sector reach carbon neutrality by 2050. The #CFMRISE program includes: ✅ A breakthrough Open Fan architecture enabling a 20% reduction in fuel consumption and CO₂ emissions compared to current engines. ✅ A design fully compatible with sustainable aviation fuels and hydrogen. ✅ Integration of electric hybridization to generate additional electric thrust and optimize engine performance.

322

1 commentaire

Seven months. One shipment — the first of many. One milestone that sets the stage for scalable growth. Earlier this month, the first physical shipment was executed through Civitanavi Systems’ new enterprise resource planning system — a major step forward for Honeywell Aerospace Technologies. In just seven months, more than 150 employees across functions and time zones came together to launch this system, which serves as our digital backbone, streamlining operations, production and customer orders. This sets Honeywell Aerospace Technologies up to scale smarter and faster. This is what digital transformation looks like in action, and it’s only the beginning!  

390

13 commentaires

𝗬𝗲𝘀𝘁𝗲𝗿𝗱𝗮𝘆, 𝗔𝘁𝗹𝗮 𝗯𝗲𝗰𝗮𝗺𝗲 #𝟭 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗼𝗳 𝘁𝗵𝗲 𝗗𝗮𝘆 𝗼𝗻 𝗣𝗿𝗼𝗱𝘂𝗰𝘁 𝗛𝘂𝗻𝘁! After tirelessly building a tool that automatically detects errors in agents, it was a huge milestone for us. We enjoyed a much anticipated bottle of champagne 🍾 But, we also received some keen questions. One that kept coming up: 𝗛𝗼𝘄 𝗱𝗼 𝗼𝘂𝗿 𝗲𝘃𝗮𝗹𝘀 𝗮𝗰𝘁𝘂𝗮𝗹𝗹𝘆 𝘄𝗼𝗿𝗸? It’s timely, because the usefulness of evals is being debated in AI circles right now. (This thread captures it well: https://lnkd.in/ewn53SC8). Are evals broken? Should we just rely on A/B tests? Is monitoring enough? From our experience building Atla, we’ve learned a few things: - Static evals inevitably go stale. The moment your agent changes, yesterday’s tests become less useful. - Manual trace review gets you to the heart of the issues, but it doesn’t scale when your agents are handling thousands of interactions. - What builders really need are 𝗮𝗱𝗮𝗽𝘁𝗶𝘃𝗲 𝗰𝗿𝗶𝘁𝗶𝗰𝘀: systems that continuously track what your agent actually does and flag anomalies. The last point is the approach Atla takes. Sashank Pisupati, PhD, our lead research engineer, wrote a blog post covering this in more detail and using an RL lens. The post explains how we’re combining production monitoring, step-level critics, and clustering recurring failures to keep evals relevant as agents evolve. Read it here: https://lnkd.in/e6XMVkrC We see this as the future of AI evaluation. Not perfect static tests. Not endless manual review. But critics that adapt as fast as your agents do. And later this week, we’ll share how these principles shaped our upgraded 𝗳𝗮𝗶𝗹𝘂𝗿𝗲 𝗽𝗮𝘁𝘁𝗲𝗿𝗻 𝘁𝗶𝗺𝗲𝗹𝗶𝗻𝗲𝘀. Stay tuned!

132

22 commentaires