hotosm/drone-tm

Related to #853 and disaster response use cases. An interesting workflow for response targeting: Often there are other aerial vehicles like helicopters around. Something like a GoPro can be strapped to the bottom with a photo interval timer. The imagery probably can't be stitched into a single ortho, but can be used for targeting response areas. Pt1: Hosting the captured photos Could go into OAM, but it mostly focuses on stitched orthos... Any way we could collate the photos into a single product, or find a simple way to link them all together without hundreds of OAM STAC items? Food for thought! Pt2: fAIr workflows for targeting Based on criteria, we could run fAIr models on the imagery. E.g. identify bridges for transport connectivity post earthquake. E.g. identify building damage to help target settlements most badly affected. This would then feed decision making for where to send resources, and in the case of DroneTM, where to fly imagery to get a better picture of the respons

E.g. Flying a city every year for change detection - every year clone the prev project. During disaster response, we want fresh imagery every few days to assess the situation on the ground: clone the same project for different dates. Common use case, simple implementation, and nice helper to those that need to do this.

Drone monitoring platform Main use case would be coordination during disaster response, but also useful for government agencies / avitation authorities to keep track of drones flying in a country. Possible requirements Where drones located How many drones imported via custom offices Who are drone pilots, who is using them Mobile version of app offline/online Arrangement with gov for liability etc This, of course, is a big risk for liability, if used by aviation authorities to plan other aerial vehicle activities based on the drone / operator locations. Drone location tracking An approach such as https://github.com/spwoodcock/nepal-pilot-dashboard WhatsApp + ChatMap --> locations of the pilots, which approximates locations of drones within a few km radius. Self reported locations Live tracking drone locations Notes from a slack channel. For that we can use the DJI v5 SDK (are they using DJI drones that are supports in the README there?). We could use this pre-built wrapper around the

I defined a no-fly zone for this project - a rectangle along the top border of the project because the project borders a military barracks to the north. Two issues: The no-fly zone is not visible on the project overview map (task selection map) When creating the flight plan there are flightplan points that enter deep into the no-fly zone that I defined. This is high risk

Specs: https://www.potensic.com/atom-3-specs.html 1/1.3-inch CMOS Quoted battery life 40 mins. More details: #863 Analysis of the files: It uses global.json and a xxx.json file for the waypoint missions, same as Atom 2. The global.json is exactly the same. The xxx.json file is slightly different: No ;[] POI suffix. Atom 2 mission files are [waypoints];[poi_list]; the Atom 3 file is a plain JSON array ending }] - no semicolon, no POI array at all, no trailing newline. The main change here. No fileName key on waypoints. Atom 2 always writes one (point_<ts>.jpg, a thumbnail for the Eve UI); Atom 3 omits the key entirely. Thumbnails are aesthetic only, so not an issue we need to action. Two enum values not previously documented: yawType: "TO_WAYPOINT" (with yaw: 0 throughout) and zoomType: "HAND". Atom 2 captures only ever showed DEFINE for both. Most likely these are mission settings the selected manually in Potensic Eve in the example provided by @zigulicd? Perhaps it means 'set yaw