RocketPy

Today at the 2026 UKSEDS - UK Students for the Exploration and Development of Space #NSSC conference taking place at #KCL, a room full of future space-sector engineers recognised how much the AI landscape in aerospace has shifted. I presented a first-of-its-kind prototype agentic Retrieval-Augmented Generation (RAG) system enabling NLP queries over rocket launches, competition results, and atmospheric conditions. The system combines vector-based semantic search for rapid retrieval, LLM orchestration for reasoning and explanation, physics-based trajectory modelling with RocketPy for predictive insights, Monte Carlo simulations and classical ML models for uncertainty and risk analysis, and integration of external environmental datasets using NASA POWER and Giovanni MERRA-2 for grounded decision-making. Launch simulations, competition database records, Monte Carlo predictions, and environmental data are stored in a vector database, allowing the LLM to retrieve relevant information and provide evidence-based answers. Classical machine learning isn’t disappearing, it’s becoming a foundational layer inside emerging AI-assisted engineering systems, opening new possibilities for faster, safer, and more informed aerospace decision-making. 🚀 🌍 💻 📈 📊 #UKSEDSNSSC #KCL

Last week, I had the opportunity to present my research titled “Predicting the Probabilistic Trajectory of Suborbital Unguided Rocket”, as part of my third-year individual project and was awarded third place in the Aerospace Engineering Category by Nicholas B.. This work examines using existing rocketry trajectory and simulation programs to develop a toolkit for trajectory and dispersion analysis that accurately predicts landing locations, including all failure modes. This is particularly useful in the UK, where amateur rockets' altitudes and launch opportunities are typically constrained by the amount of available launch space with landowner permissions. This work aims to reduce these constraints by better understanding rocket dispersion patterns, recommending optimal launch inclination & heading, and reducing uncertainty around low-altitude winds through experimental methods. This work has already been applied during the design stages of ManSEDS (Manchester Students for the Exploration and Development of Space) upcoming climate-sensing high-altitude rocket (launching to around 21000 ft), as part of the Launch4Change program. This helps improve the safety case and reduce dispersion uncertainty in a vehicle intended to launch beyond the typical altitude limit of 16000 ft. This work has also been applied to develop the safety case, alongside the existing UKRA safety code, for an outreach event within the wider Manchester area to support STEM skills and promote climate awareness as part of the Launch4Change programme and UKROC competition. Thanks to my supervisor, Kate Smith, for supporting this project. I’m excited to keep pushing forward with this research, developing recommendations and a toolkit on top of RocketPy that rocketry teams can use to improve safety cases and enable higher altitude launches within the UK. #UoM #ManSEDS #Launch4Change #UKROC #AerospaceEngineering #RocketPy #UKRA #UKROC

Workshop sign-ups for NSSC 2026 are still open! 🚀 From debating the ethics of human spaceflight to running AI-powered launch simulations with RocketPy, our workshops are your chance to challenge ideas, build practical skills, and connect directly with the space industry. This year’s workshops include: - Space Debates - AI-Powered Aerospace Simulation & Launch Planning - CV Workshop - Project Management - KISPE Satellite Practical Workshop - Airbus Early Careers Workshop Each workshop has its own registration link (free for NSSC delegates!), so make sure to head over to our website and secure your spot. Browse all workshops here: hub.ukseds.org/events With tickets for the workshops nearly sold out, now’s the time to sign up! See you in London in very soon 👀 #NSSC2026 #UKSEDS

Safe launches start with accurate data. 🌤️🚀 We are proud to support RocketPy at the European Rocketry Challenge (EuRoC). By utilizing Meteomatics' API, RocketPy simulates rocket trajectories using real-time, high-resolution weather data, ensuring that every launch is as safe as it is spectacular. See how our weather data supports the next generation of aerospace engineers: https://lnkd.in/eVrq_69v