Profile
Blog
Photos
Videos
Arrival
What a neat and tidy place. Sure there are wiggly streets going here and there but they all have a sense of being cared for despite that its one of the world's biggest metropolises. I was very excited when I arrived, experiencing what the city and people were like. I arrived very late at night and caught the monorail from the airport to another station where I transferred to normal light rail. The monorail was very quick, much better than the rattling slow one in Sydney. I think I saw many of the infamous Japanese company men on the train… that they looked like they had just finished work (at midnight) although I'm not sure if this observation is due to my having developed an overblown sense of respect for Japanese men’s work ethic or because they were in fact late home from work.
There was a very helpful airport staff member who helped me catch the monorail from the airport, it was nice that they employed someone for the purpose of helping clueless foreigners. I’m glad he showed me how to catch the train as otherwise I probably would have tried to get a taxi… and paid the price. I was to learn the following day that a fellow conference member had paid $300 to go to the same hotel as me, a 40 minute taxi ride!!!!!! And I had been shocked to have to pay $50 to go 9km when the trains stopped at 1am, halfway to my hotel…
Day 1
Thursday, 6 September 2012
Today was the first day of State of the Map 2012, a conference about OpenStreetMap at Tokyo University. The day was about disaster response, given the Fukushima disaster during the Japanese earthquake last year and the earthquakes in New Zealand. There were a number of Japanese presenters, describing how they’d contributed to or used OSM as part of the relief efforts. The impact was still strongly felt for one presenter who ended his presentation in tears, describing how, when on a mapping trip, he’d come across the body of a woman who’d not survived the tsunami.
The presentation about the New Zealand quake was about a very thorough effort to map the temporary state of places like Christchurch after the disaster, so people would have a clear understanding of the present state of things. He regularly talked about using a '=temporary' tag on features, which I thought would have required a lot of work for upkeep because someone would then have to go and discover the status of such features and re evaluate whether they should still be tagged temporary.
An issue discussed after his talk was the negative impact of out of date satellite imagery in such situations where 'armchair mappers’ can actually detract from a response effort by digitisting features that don’t exist as the imagery is older than the disaster event. One conference member raised the possibility of a layer in JOSM which would alert such people to the parts of the image which were no longer valid. I responded by saying that perhaps we needed to expand the scope of open source mapping to also include aerial imagery. The first responder mentioned openaerialmap as one option, although this is now defunct. I think my suggestion has validity because UAVs are very cheap now (Swinglet Cam was the example I gave, being $5000) and they are easy to operate (autopilot software is included), returning usable imagery and a DEM in just half an hour! I think a similar massive global collaboration needs to take place with imagery. Then the crowdsourced approach to reference spatial data will be complete as such data would be expressed both in vector and raster format. OSM has the vectors covered, now we need the rasters.
I also met a great many enthusiastic people with similar interests to me. I spent half an hour talking to a mathematician from Poland who was studying to be a pilot, but was employed as a programmer who did some mapping work. We talked a lot about remote sensing and how it is now within reach for people to launch their own satellites through NASA’s cube sat program. I also met 3 people from Skobbler, including the boss. I met three people from Mapbox. At the start of the conference there were some brief words from the founder of OSM, whose name escapes me. I met Kate Chapman from the Humanitarian OSM Team (HOT) who interviewed me for a position in Indonesia although the position ended up being cancelled. People also mentioned that the developer of OSRM was at the conference, so I will have to show him the moto trip route I’ve planned using it and get a photo!
All in all a fantastic day, so glad to be here. Whilst I was watching speakers I made a map of the 10 urban poor communities for which we’ve uploaded house footprints. This was done by creating 11 data frames all with OSM tiles via the import basemap function. This made the map very slow, but I got there in the end. I have a key map of Phnom Penh and another of Cambodia, also using OSM tiles. It looks good and I’ll show it in my lightning talk tomorrow.
Day 2
Friday, 7th September 2012
Today I met many cool people! I got a picture with Dennis from OSRM, and a picture with the Skobbler team (well, 3 of them). I also got to talk to Steve, the guy who started OSM! When I asked how long ago, he said 8 years and 2 weeks… That precision says something. He has a surprisingly boyish face, but is very tall. He seems to be on the edge of being revered by most conference attendees. Everyone was very keen to talk to him and I asked him if the fact that some companies were now paying some of their staff to contribute to OSM full time would change the nature of the project. He said it is not the way of the future, it has been the way of the past and present for a while now. I didn’t really get anything out of him about whether he thought this would mean contributions of mere unafilliated volunteers might then be viewed differently. I want to get a photo with him as well, tomorrow though as I was shy today.
We had many interesting tech presentations today and I met Falko Schmidt, a fellow supervisee of Stephan Winter! He was from Bremen University (and involved with SFBTR8), so before Stephan came to Melbourne. We traded names for a while and seemed to know many of the same people such as Martin Raubal and Kai-Florian Richter. I said to him that I’d been waiting for the term spatial cognition to come up during conversation at the conference… It was great to have a chat and again find yet another person I had a lot in common with.
We had some interesting presentations but the best was Dennis’ one about OSRM and also one from the guys at Bosch, who forced them to cut off the livestream for theirs’. The OSRM one was so interesting because I’d been using it for a number of months and it was a great way to promote OSM to people who mainly use online maps for routing. Dennis is also a really cool guy and I was able to introduce him to Math via email, who had told me about OSRM originally. The coolest thing to come out of the presentation was that OSRM may now be included in the main OSM website so that people can use it for routing! This means that OSM will finally, in the eyes of the public, begin to offer a viable crowdsourced navitation alternative.
The Bosch presentation I think was cut from the livestream because it involved revealing some proprietary heuristics they’d developed to simplify complex intersections (cloverleaves etc) in the road network of OSM for their routing software. Bosch currently provides navigation solutions for car manufacturers but they come with Navtech data, so this is part of their effort to bring costs down by switching to OSM. Complex intersections simplification was done to speed up computations for long routes which took in many such intersections. Such an item or feature might be made up of an on ramp, another section of the route within the intersection followed by a part exiting the intersection. At certain zoom levels this level of complexity is unnecessary, especially for map visualisation where a cloverleaf overpass, for example, is not worth showing. In these cases, they computed only simple crossways intersections, reducing the number of ways that needed analysing. Additionally, they seemed to present a heuristic for simplifying route calculations that went between cities on major highways that took in major intersections, showing how it would only evaluate route alternatives within the highway=primary or above tags, reducing the number of alternatives greatly given that most intersecting roads would have been less than this minimal tag level. At least I think that’s what they were talking about. I’m no software engineer or algorithm monkey.
The day ended with a nice river cruise and then many of us got together and braved the labyrinth called the Tokyo Subway System on the way to a stop called Roppongi. This was an area with many bars, although they seemed to cater to westerners and were kind of lame. Nevertheless, the one we went to had Microsoft Surface tables, essentially 50" or so touch screens! They were notoriously difficult to use as the touch interface was crap, but one cool program was a Bing version of Google Earth, so we had fun sharing where we lived and checking out other cool places on the map.
Oh! I nearly forgot, my lightning talk about the 3 Cambodia OSM Mapping Parties happened! It went very well! I had 2 questions, one guy wondering how many Khmers were still mapping. I’m not actually sure, perhaps I could ask on the PPMM fb page, but I emphasised that I was focused on giving them skills as I was just a foreigner who knew little about the content that should go on the map. Another guy came up to me afterward and wanted to find out more about mapping in Cambodia. I’ll be having a chat with him tomorrow. So, all up, a great day!
Day 3
Saturday, 8th September 2012
Today was the last day. I got some more photos with 'famous people' ;-). They were the ITOworld team and the Mapbox team. It was really great to meet people from these companies where I've only ever seen their website. The theme of today was 'community' although to be honest it was not what I expected. We did have a great presentation by Kinya Inoue called Mapping of historical places in Japan. This was really interesting because he had made so many fantastic contributions to the map (two here). He had some great stories to tell like getting GPS tracks at a castle but accidentally walking in the line of fire at an archery range... Another great presentation was by Tim Waters, who gave many funny observations about what it's like to be a compulsive OSM mapper. We could all relate to that. Another great bit of his presentation was about a questionnaire to determine your OSM purity. Falko Schmid, who I mentioned yesterday, also gave a presenation about an attempt to help people of Laos through an ICT4D project. This was to create MapIT, a photogrammetry program to automatically digitise small features for the map which are smaller than the circular error of a hand held GPS. His main example was a set of adjacent car park spaces at his university, with there being a
- comments