#EvacPaperAlert: EvacUnet: A deep learning-based crowd density prediction method for assisting #evacuation analysis Just published on #FireSafetyJournal 🚧 #toll_access paper (subscription or payment required). 🚧 🔗 Link: https://lnkd.in/gRXB4pd9 🚨 A deep‑learning solution called EvacUnet is being shared, which turns architectural floor plans into instant, high‑resolution pedestrian‑density maps for any moment during an evacuation. 🔹 Why it matters – Fast, accurate density visualizations help safety managers make real‑time decisions, improving the safe egress of occupants.   🔹 Robust training – Built on a large‑scale, multi‑building evacuation dataset, EvacUnet works across diverse structures and occupancy patterns.   🔹 Consistent performance – Even in highly complex scenes, the model maintains a stable, rapid inference speed.   🔹 Cost‑saving potential – Engineers can now run risk‑analysis for intricate buildings without the usual time‑ and money‑intensive simulations.  This technology paves the way for smarter, more efficient emergency‑response planning in the built environment.  💡 Stay Connected: Don't miss out on any updates! Follow us on #X for more insights and exclusive content: https://lnkd.in/guNVuUfN #SafetyScience #DeepLearning #EvacUnet #EvacuationSafety #RiskAnalysis #AI #BuildingSafety #EmergencyResponse #PersonnelManagement 🚀🏢📊🧠⚡️

Smidnya Technologies | AI Vision for Earth’s Early Warnings The future sees before it feels. At Smidnya Technologies, we’re reimagining AI Vision Inspection to go beyond production lines — into the heart of the planet itself. Our intelligent systems now interpret satellite and aerial imagery to identify early signs of floods, wildfires, and landslides, turning raw data into life-saving foresight. This is where deep tech meets deep care — AI that doesn’t just detect defects, but protects what matters most. 🔹 Early disaster prediction powered by vision AI 🔹 Real-time environmental monitoring 🔹 Scalable intelligence for a safer world Innovation, for humanity. That’s the Smidnya way. know more: https://zurl.co/jOLx1 #SmidnyaTechnologies #AIVision #DisasterDetection #AIforGood #SmartEarth #FutureOfAI #Sustainability #Automation #DeepTech

🚀 𝗠𝗮𝗸𝗶𝗻𝗴 𝗦𝗮𝘁𝗲𝗹𝗹𝗶𝘁𝗲 𝗗𝗮𝘁𝗮 𝗪𝗼𝗿𝗸 𝗳𝗼𝗿 𝗛𝘂𝗺𝗮𝗻𝗶𝘁𝗮𝗿𝗶𝗮𝗻 𝗔𝗰𝘁𝗶𝗼𝗻 🛰️ Honoured to share a short, 15-minute dialogue I had with Bill Greer in the Humanitarian AI 𝘛𝘰𝘥𝘢𝘺 – 𝘝𝘰𝘪𝘤𝘦𝘴 mini-series (organised by Brent O. Phillips). As geospatial practitioners, we are recognising more than ever the critical need for humanitarian actors to have reliable, independent, and affordable access to satellite data. In our conversation, we emphasised the importance of initiatives like Common Space and how this open satellite constellation and data platform can support life-saving decisions during crises, ensuring access to imagery is not influenced by commercial priorities or limited availability. We also explored the need for 𝗻𝗲𝘄 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗺𝗼𝗱𝗲𝗹𝘀 for data sharing to encourage 𝗰𝗼𝗹𝗹𝗮𝗯𝗼𝗿𝗮𝘁𝗶𝗼𝗻 𝗮𝗻𝗱 𝗶𝗻𝗻𝗼𝘃𝗮𝘁𝗶𝗼𝗻 as public funding decreases and the commercial imagery market evolves. While sensors are becoming more capable and satellites getting smaller and more affordable, long-standing government programs like 𝗟𝗮𝗻𝗱𝘀𝗮𝘁 (from 𝗡𝗔𝗦𝗔) are gradually being phased out, and 𝗘𝗦𝗔 is shifting towards a more research-focused approach with its 𝗙𝘂𝘁𝘂𝗿𝗲𝗘𝗢 program. This transition means these organisations are unlikely to provide the future coverage humanitarian actors require. Initiatives like 𝗖𝗼𝗺𝗺𝗼𝗻 𝗦𝗽𝗮𝗰𝗲 are therefore essential for bridging this emerging gap and playing a key role in shaping the future of humanitarian action today. 🎙️ If you’re interested in the intersection of AI, earth observation, and humanitarian action, you can listen to the episode here: 🔗 https://lnkd.in/ejuvKifh #HumanitarianAI #EarthObservation #SatelliteData #OpenData #HumanitarianInnovation #CommonSpace

Smidnya Technologies | AI Vision for Earth’s Early Warnings The future sees before it feels. At Smidnya Technologies, we’re reimagining AI Vision Inspection to go beyond production lines — into the heart of the planet itself. Our intelligent systems now interpret satellite and aerial imagery to identify early signs of floods, wildfires, and landslides, turning raw data into life-saving foresight. This is where deep tech meets deep care — AI that doesn’t just detect defects, but protects what matters most. 🔹 Early disaster prediction powered by vision AI 🔹 Real-time environmental monitoring 🔹 Scalable intelligence for a safer world Innovation, for humanity. That’s the Smidnya way. know more: https://zurl.co/weTBs #SmidnyaTechnologies #AIVision #DisasterDetection #AIforGood #SmartEarth #FutureOfAI #Sustainability #Automation #DeepTech

Stop Guessing Which Pipe Will Explode Next (The Tech That Saves Cities Millions) 💰🛠️ We’ve all seen the mess: a massive water main break, traffic gridlock, and the city scrambling to fix it. It’s costly, disruptive, and honestly, a huge embarrassment. For years, cities just reacted, fixing the oldest pipes and hoping for the best. That era is over. I want to talk about the real game-changer in asset management: GeoAI using Graph Neural Networks (GNNs). This isn't just "smart mapping." It’s about building a digital twin of our physical reality. The Breakthrough: Seeing the Network's Stress Think of a city's pipeline system like a complex human body. 1. The Old Way: We used to just look at a pipe's age (like a person's birth certificate) and guess if it was sick. 2. The GNN Way: We model the entire system as an interconnected graph, letting us factor in all the real, messy variables: • Soil Type: Is that pipe sitting in corrosive soil that eats away at the metal? • Traffic Load: Is it under a street where a thousand heavy trucks roll by every day? • Network Strain: Crucially, the AI models how the high pressure from a new pump station right here is stressing an old pipe segment just down the line. The GNN explicitly models this cascading stress. It identifies that one specific junction—that single weak link—that is mathematically guaranteed to fail first. The Result? Utility teams stop wasting money on random pipe replacements. They get a map showing the 1% of the system that needs immediate attention. That’s the power of predicting disaster, not just cleaning up after it. If your company owns a network—any network!—you should be asking your tech team about GNNs. It’s the difference between hoping for resilience and engineering it. Thoughts on where this tech could be applied next? Power grids? Rail lines? Drop your ideas below! 👇 #Infrastructure #GeoAI #GIS #AssetManagement #GNNs #SmartCities #Innovation

I argue there are few better places than #TEDAI to see firsthand that the promise of AI sits square at the intersection of humans and technology. I end the week in awe of the incredible minds who shared their wisdom at #TEDAISF at the beautiful #HerbstTheatre who are leveraging this technology to elevate humanity. Dr. Nathan Price of #Thorne shared how his organization leverages AI and data to navigate biological complexities to offer "personalized, scientific wellness" - using technology to listen to the cellular whispers of the body to understand the early warning signs for the major human diseases so they can predict and prevent them. Philip Johnston at #Starcloud was the embodiment of this years TED AI theme of, "Are We Bold Enough?" by proving that, against all odds, that what most believed impossible was anything but. Starcloud-2, what will be the first-ever data center in space, will be fully operational in sun-synchronous orbit by 2026. "...a recent study by the European Commission concluded that orbital data centers will significantly reduce greenhouse gas emissions from grid electricity and eliminate freshwater usage for cooling." https://lnkd.in/gkZPrU2e Kicker? I even got to hang out with a group of #robots showcasing the #artofpossible at the intersection of hardware & software. ☺️🤖

AI is transforming the way we design, operate, and manage satellite systems and offer services to our customers - and its true impact on the space sector is only just beginning. I recently joined an industry roundtable with Via Satellite to discuss how AI is helping us at Eutelsat to drive efficiency, resilience, and innovation across our global networks. From predictive maintenance and customer insights to autonomous fleet management, we are already seeing tangible results that improve both our customers’ experience, the services we offer to our customers and our operational processes. In the interview, I also shared a preview of what might be coming next: AI-driven automation and agentic systems that can make real-time, data-led decisions across multi-orbit constellations like our GEO-LEO network. This will be key to unlocking the next phase of scalability and sustainability in space. AI is set to enable a new generation of “zero-touch” satellite operations - where networks can optimise themselves dynamically based on demand, ensuring greater safety, responsiveness, and environmental responsibility. You can read the full roundtable discussion on Via Satellite’s website: https://lnkd.in/dKtTiyWN

Great insights on how AI is transforming satellite operations at Eutelsat! Don’t miss the article where my manager Mariam Kaynia shares our vision at the Via Satellite roundtable on building scalable, sustainable and zero-touch networks. #Eutelsat #AI #Satellite #Innovation #Automation #SpaceTech

ODOT's Bridge to the Future: Artificial Intelligence (AI), Machine Learning (ML) & Geographic Information System (GIS) for Scalable Asset Data Analytics Excited to be speaking at #IHEEP2025 to showcase how the Ohio Department of Transportation (@ODOT) is transforming transportation asset management through cutting-edge technology. ODOT is leveraging AI, ML, computer vision, GIS, and reality data to automate and enhance bridge clearance analysis across ~3,000 bridges statewide. By tapping into existing mobile LiDAR and photolog imagery, we're generating accurate vertical clearance measurements. These efforts aim to reduce manual data collection, improve asset inventory accuracy, and support multi-modal planning and investment strategies. In future phases, ODOT plans to expand the scope to include other facility attributes. This initiative is a game-changer for asset management, GIS, reality capture, and AI/ML applications in large scale infrastructure. If you're attending IHEEP, join us to explore how you can benefit from similar innovations. #IHEEP2025 #ODOT #AtkinsRealis

🌍 Satellites + AI aren't just for high-tech labs. They can be a savior for communities with limited resources. By analyzing neighborhoods from above, we can inform zoning, green space planning, and health service delivery decisions without the need for costly field surveys. This means even under-resourced areas can access the same cutting-edge insights to build healthier, more equitable futures. 👉 Curious how it works in practice? Check out the full video here: https://lnkd.in/dNc5Bkmi #AI #GeospatialAI #HealthEquity #EarthObservationForHealth World Health Organization, Centers for Disease Control and Prevention, European Space Agency - ESA