Hi Want to give some background and structure to the proposed OAM design workshop.
Camp Roberts. Last week, David Bitner, Chris Lippitt, and a bunch of other great folks participated in RELIEF experiements coordinated by Star-Tides [http://star-tides.net/] and the Naval Postgraduate School [http://www.nps.edu/]. We went to test software integrations and simulate field conditions that mapping projects for humanitarian relief would face in places like Afghanistan. Most of the groups involved, like myself representing OpenStreetMap, blogged about the awesome work there, recommended reading... http://delicious.com/tag/camproberts The main driver for the week was the NGA authorization to share sub-meter resolution imagery of eastern Aghanistan with "civilian partners". Google brought down a Fusion server, and had that hefty server producing tiles. That job ran overnight and saved tiles to a portable drive that is now sitting in the only bar in eastern Afghanistan. Though this is very very awesome, we lamented that this was the only resulting copy of this imagery, and the uplink from that bar is not conducive to downloading hundreds of gigabytes of data. We naturally started talking about OpenAerialMap. OpenStreetMap had an excellent showing last week, providing data and integrating in multiple ways .. why wasn't the same concept working for aerial imagery despite everyone's herculian efforts over the last couple years? Well we all know the story here thanks to Chris's write up on the real world experience on his awesome app, Cristiano's workshop at SOTM, and the thorough threads from yesterday. In my view, basically OSM could succeed because its resource curve was not steep, allowing for a very small group of folks to build something usable, and giving space for a community to form, step up and offer resources as needed. OAM is not in this position. The processing and storage requirements of rasters versus vectors means that, from the start, a great deal of resources need to be on hand OR a great deal of cooperation needs to be in place. We have tried to coordinate OAM over the list and wiki for the past couple years. There has been much exploration and hacking on ideas. Yet, we remain today without a concrete steps forward. After the quick work at Camp Roberts, it's clear that what OAM has lacked is a clearly articulated set of use cases and plan to drive the project. We feel that face-to-face time among interested parties would help clarify these use cases, and form a plan for moving forward. MONTERREY When the sponsors of the Camp Roberts exercises, the Naval Postgraduate School and Star-Tides, heard this story, they were intriguied and now want to help. They are looking into the possibility of hosting a 3 day design workshop at NPS, and possibly covering some travel costs. This would be a very generous offer that could set OAM in the right direction. The goal of the workshop would be to produce an achievable roadmap for the development of OAM. There seems to be general consensus already that some kind of federated structure is the way forward. Many details within that need to be worked out. In the time preceeding the workshop, we'd want to debate and capture as many ideas as possible. I heartily second David's call to start feeding discussion into wiki pages, and Richard's call to work towards a meeting agenda. In my mind, here's a few of the general types of things we want to capture, perhaps each on a seperate page of the wiki. This could be something like a structured brainstorming, with the pros and cons of all ideas laid out, an attempt to catalog the thinking of everyone in the community. * Use and Problem Cases. To drive the design process. We've seen everything from someone wanting to share photos from their DIYUAV, to small municipalities posting their imagery, to complete global coverage. * Goals. What are the potential goals. * Current Data Sources. Potential Data Sources. Who is offering, where is the data, what is the format. * Licensing. Imagery is available in many licenses. What are possible ways to deal with this complexity? * Architecture. What does a federated imagery sharing system look like? Perhaps we could divide this up into a number of different scenarios, that cover a broad number of ideas .. everything from a very simple catalog, to blue sky dreams. * Standards. What's out there and what's missing? * Tools and potential code. What are our building blocks? * Potential Hardware and Hosting. Even in a distributed model, someone needs to offer up resources for this thing to run. * Interested individuals. Your organization and general take on OAM. Would anyone be ready to step up and take this, or some other suitable structure, and create stubs, etc in the OAM wiki?? In this wiki brainstorming, we can remain open to what OAM actually is. In the most general sense, many folks want a way to easily share and combine aerial imagery. Once we get to the workshop stage, the driving goal will be to take the collected ideas through consensus into a workable design roadmap. What is achievable in the next six months? What ideas are years away and how do we possibly get there? If we make quick work of the roadmap, there may also be time for some quick prototyping and hacking. Components of the roadmap could possibly be developed into fundable projects, and acts as motivation towards more cooperative imagery licensing within the US government. PRACTICALITIES At this moment, we're still waiting for specifics from NPS. We'd like to see a quorum of core interested people attend in person. Myself and John Crowley from Star-Tides would like to take a neutral faciliation role to help the process. For those that can't attend, we would make every effort to include remote participants, through IRC or Skype. The results need to be acceptable to the OAM community as a whole. Perhaps these questions can be briefly answered by everyone in the wiki. Who? Who's interested to attend. There's a pratical limit to the size of the workshop. Want to make sure the workshop is representative. Thinking that a very brief application process might be called for. When? Over the next couple months, which weekends are possible for the greatest number of these people. Sometime in last half of September, first half of October. Where? NPS is in Monterrey. Realize that's more difficult for some than others, but that will be the case for this globe spanning project. There will potentially be support in place. OK! Eager to hear your thoughts on this plan, who and when, and most especially seeing the discussion translated into the wiki. Cheers Mikel
_______________________________________________ talk mailing list [email protected] http://openaerialmap.org/mailman/listinfo/talk_openaerialmap.org
