I should turn my comment under this LinkedIn post into a post: https://www.linkedin.com/posts/eliot-higgins-955822216_when-satellite-imagery-goes-dark-new-tool-activity-7447300235669852161-ZcSX Blog Post: https://www.bellingcat.com/resources/2026/04/07/tool-damage-assessment-destruction-sentinel-satellite-imagery-iran-us-gulf/ Summary: The Bellingcat article introduces the Iran Conflict Damage Proxy Map, a new open-source tool developed by Ollie Ballinger to estimate building destruction across Iran and the wider Gulf region amidst ongoing conflict and internet blackouts. As commercial satellite providers like Planet Labs and Maxar increasingly restrict or delay high-resolution imagery in the Middle East to prevent its use for "Battle Damage Assessment," this tool utilizes publicly available, medium-resolution Sentinel-1 radar data to fill the information gap. Unlike optical imagery, which can be obscured by clouds or intentional blackouts, the radar-based approach detects physical changes in structures by comparing pre- and post-event satellite passes, providing a vital resource for journalists and researchers to verify the scale of military impacts in areas where on-the-ground reporting is restricted. **My comment**: Sentinel-1 is an excellent platform for this use case; the blast impact sites are (regrettable) quite large, so even after a 5-day revisit and clean up, the scars can still be seen. Unfortunately, this also highlights the primary shortcoming: while it can make observations under almost any atmospheric condition, the weekly frequency is a real limitation for real-time monitoring and on the ground humanitarian support.

Hi @Dawid Hanak, I reread your post and realised I overlooked the 'problem I tackled' part. To be honest, I haven't fully solved this yet, but I'm working on it! My comment reflects a challenge I’m currently facing with the Common Space group. We’re addressing these exact gaps, but since our solution isn't operational yet, the team isn't quite ready to share the research or methods. Do you think there’s value in sharing the 'problem' stage before the 'solution' is complete, or is it better to wait until we have concrete results?

@Dawid Hanak - there might have been a better thread to drop this.

The post is "live" https://www.linkedin.com/posts/gvddool_earthobservation-remotesensing-machinelearning-activity-7442591207559462912-bZXM

@Dawid Hanak - That was perhaps the only good thing about the lockdowns: being forced to do nothing or being limited in what you could do! I’m still baking my own bread (when I’m in a larger kitchen) as a result, and there’s nothing better than spending 30 minutes doing something completely different, away from the screen!

@Dawid Hanak - Whether you have a bread maker or not, there's nothing quite like home-baked bread! We don't have one, but I found an amazing recipe that's almost the same as having one: it uses warm and cold fermentations, making it super easy — you just pop it in the oven in the morning, and enjoy fresh bread for breakfast or lunch ;-) Hope "doing nothing" will become popular - yesterday I saw a short video on LinkedIn about a lizard, and the OP said the same - he watched the creature doing its thing for 30 minutes, nothing more.

Interested in following the discussion, AI and Ethics is an important topic, and not widely, or often enough, discussed

On a related topic - the following dropped in my inbox: Humanity in the AI Era” Webinar – 18 March 2026 The current era of rapid AI evolution presents both promising benefits and complex risks. Global conversations increasingly recognize ethics as an essential foundation for successful AI adoption, affirming a humanity-centered approach for collective betterment. This talk explores what this means in practice, reflecting on the broader implications of responsible AI, how technological trajectories can shape societies in materially different ways, and how these ideas connect to widely shared global goals, drawing on some recent examples. Participants will also be encouraged to think through some forward-looking action steps that could help individuals and institutions to better adopt AI responsibly and harness the opportunities this moment presents. The discussion is structured to be accessible for participants across digital literacy levels. * This event will be recorded https://zoom.us/webinar/register/WN_rDoI3qaeTXmggYFpY-uMJQ#/registration

It's true that consistency helps build trust and attract subscribers, but sticking to a schedule can be hard. Aiming for a weekly or bi-weekly routine over the 6-12 months might be doable. However, for me, the hardest part is finding topics I care about enough to set aside time to write 1000-5000 words each time.

@Dawid Hanak - every month (or so) is a nice cadence. It leaves enough time between issues to do a proper deep dive. Some of the more frequent newsletters I see on LinkedIn are either summary-style collections of articles that don’t add much insight beyond what’s already in the links, or quickly generated AI slop aimed at clickbait and self-promotion.