Where 2.0 Conference

Day 3 begins with Rio enjoying the early morning sun rays.

After an exhausting 25 sessions followed by the WhereFair, I'm ready to relax and take it easy today. This year, the conference is held in Burlingame, California in the Marriott Hotel. It is right in front of a marshland facing the runway of the San Francisco airport, and so, when you decide to step outside, you hear the constant deafening roar of the airplanes. Aside from the noise, it is a beautiful and peaceful location.

Going back inside, the day begins pretty much where it ended yesterday. Day 3 features another monster marathon of an additional 22 sessions. By the time lunch rolls along, I think that even I, a self-proclaimed Geo-Geek, feels like my geo-quota has reached its limit. I mean, how much geo-talk can you take? 40+ sessions in 2 days?? This quantity over quality approach has pretty much numbed many of the particpants, and the crowd has slowly dwindled as the day progresses. Nevertheless, there are a few highlights of the day. Lior Ron, from Google Map's team, announces a brand new feature on Google Maps: the "more" button, which adds Wikepedia material as well as photo's on the map.

Another presenter that caught my interest was Erik Hersman, a self-proclaimed activist who has built a site called Ushahidi that tracks activities happening on the ground in troubled states (in his example, it was Kenya's unrest following the recent election). What piqued my interest was the Ushahidi incorporated a Simile timeline that synced nicely with a Google Map. Simple and effective way to mash time and space elements.

So, after sitting through 40 geo-spatial sessions, I am more than ready to geo-locate myself as far as possible from here. There is just so much "where" that one can take. Oh, but I sure can't do it without grabbing some of the snacks...

At the end of the 3 days, I am struck by a nagging thought about all these geo-spatial technologies and direction this field is heading into. Granted, many, if not most, of the presenters were small start-up companies trying to vie for a spot in the saturated geo-industry, and at times it felt more like an audition session for venture capital. Kind of like a Geo-Spatial Idol. A large majority of the content appeared to revolve around the notion that "we want to know where YOU are", because with that information, innovators can invent all sorts of information networks: you can know where you are, you can find out where your friends are, you can find out what the inside of a restaurant looks like without going there, you can have a machine tell you what is around you without even looking. After some time, I am thinking, what are we doing with ourselves? Are we drone-atizing our existense? Are we all really willing to so easily give away our geo-location to others? Is this really where geo-spatial technologies should be heading?

At the airport, on my way home from the conference, I ran into a friend and fellow participant, Bruce Raup from Boulder, Colorado. He summed the entire conference for me. "They are solving problems I don't have".

I for one, would like to keep on geo-tagging, all the while making sure that I don't get geo-tagged.

Of course, day 2 begins early in the morning with Rio serving as my alarm clock. He let me know that I needed to be sharp and ready for a long day ahead of me.

The conference kicked off in earnest today with a marathon session of 10 minute presentations. All in all, the day had a total of 25 back-to-back speakers, each making their claim in this increasingly crowded geo-spatial world. The diverse array of speakers certainly made for an entertaining forray into the eclectic nature of the booming spatial industry. Speakers represented big companies (Google, ESRI, Microsoft, Yahoo, Nokia, Autodesk), small start-ups, and one academic (no, it wasn't one of us). The morning highlight came during John Hanke's (founder of Keyhole) presentation. He got up on stage and talked about a mystery partner that Google now had in terms of enhancing their geo-data coverage. To the surprise of many (including myself), no other than Jack Dangermond (founder of ESRI) got up on stage! It was certainly odd to see Google and ESRI talk on stage as though they were long time friends.

It is further compounded by the fact that to my knowledge, ESRI had been working hard in their 3D endeavors to compete against Google Earth, primarily with their ArcGlobe product. Their announcement to thus allow a large part of their GeoWeb data plan to merge with Google Earth was a surprise, and perhaps a surrender in some aspects. It is also interesting to note that ESRI's newest 9.3 release of their famed ArcGIS suite has a new commitment to openly integrate their services with Google Maps and Virtual Earth, something they had not supported to in the past.

The afternoon show was dominated by Vincent Tao, director of Microsoft's Virtual Earth team.

Vincent talked about Microsoft's focus for more "real" 3D environments by stressing their priority to create highly immersive 3D visualizations using their 3D geo-browser. He also demonstrated the direct tie-in between their famous "bird's eye view imagery" and their 3D models, switching back and forth in real time between the two different visualizations. He also demonstrated a project that they will be releasing soon that showed how regular photographs can be virtually geo-referenced and located in the 3D model, in the correct angle and correct orientation.

Other interesting talks included a human geography/crowd simulation project (http://geosimulation.org) by Paul Torrens, a professor from the University of Arizona.

A crowd streams through a dense urban setting

On a side note, one thing that fascinated me throughout the day was just how completely techno-geeky the conference was! 2 in 3 participants (including myself), had their laptops open during the presentations, making for a nice firefly effect in the dark conference hallroom.

Finally, at 6pm, all 25 presentations were finally over. The "Where Fair" was to follow at 7pm, where 15 groups (including our very own Hypercities project which I was to present) would showcase their projects in a county fair type setting. I was fortunate enough to get put in a booth with a plasma screen (only 4 groups were so lucky!), but instantly felt intimidated when my neighbor began setting up a project called "celestial", where he demonstrated a 3D browser that featured, oh, the entire UNIVERSE!

So, feeling rather micro in comparison to the universe, I nevertheless set up my booth to showcase ancient Rome in Google Earth on the plasma, and Hypercities on Google Maps on the monitor. Despite the fact that I was out-crowded 10-1 by my celestial neighbor, I never had a solitary moment in the 2 hour exhibition, and met some very interesting people throughout the night. People from Holland, Korea, Japan, Singapore, Malaysia. Academics from Stanford, Berkeley and Harvard. And all sorts of industry folks stopped by to marvel in Ancient Rome and travel back in time in Berlin and Los Angeles. What made the Where Fair colorful was that it coincided with what seemed like "happy hour", and many a "happy" people stopped by my booth which was strategically located next to the bar.

So, after an exhausting day, I finally have some time to reflect. What seems to be obvious is that there is a lot of jostling by the small companies to get into a niche and survive, while there is definitely a lot of rivalry and some healthy collaborations happening from the big players. In the end, I am happy to be in academia and be left out of all the stress of industrial survival, sit back, relax and hope that good technologies are born through this democratic business competition. When the fittest survive, we are sure to benefit from some incredible innovations happening in the geo-spatial realm, as evident from an impressive array of ideas and creativity by the bright minds at this conference.

The first thing to do when you get to a conference is to settle down in your hotel. Since I've brought my family over to Burlingame, CA, my son Rio spent little time getting comfortable in our hotel room...

After getting registered for the conference, I was greeted by a nice Plasma screen TV that displayed a Google Maps mash-up of all the participants. It automatically panned and cycled through each registrant's location. Nice touch... I kept waiting for my name to show up, but soon realized that it could take forever to get to me, given that there are close to 1000 people here...

The first day of the conference (Monday) is about workshops. These "hands-on" tutorials are run by the different industry leaders. The first workshop I attended was hosted by the Google Maps API team. In order to make everyone feel very "googley-ish", they brought their office furniture with them, and adorned the room with their colors.

The first part of the presentation was about the history of Google Maps, example of sites that use the API, and the future direction. If you read the Google Maps API blog religiously like I do, there was nothing new here.

The second part was a rapid-fire, hands on tutorial led by Pamela Fox, describing the process to develop a pizza store locator using PHP and MySQL. Frankly, I thought she lost half the audience in the first minute, but at least she kept referring to pizza searches which kept us all in tune since it was just before lunch time. Here's her presentation outline:

http://imagine-it.org/storelocator/

However, the real fun part of the session was that I was sitting directly behind the entire Google Maps API team. During each break, I was able to chat with a couple of them. I met their documentation leader, Tom Manschrek, Tech Leader Ben Appleton, and Lead Developer Pamela Fox. When I described our Hypercities project to them, they showed great interest, and so I sat them down and gave them a demo of the site. Tom, being a History major, and having recently traveled to Berlin, seemed thrilled with the idea of space and time travel. After showing the project, I described the importance of the temporal dimension. Interestingly, although they all agreed that this was an inherent paradigm of the way we live (in space and time), they did not reveal any hints that Google Maps is making any headway into this realm. I insisted that this needed to be a priority in the API, especially given that this is already a built in function within the Google Earth/KML environment. Nevertheless, there was something very satisfying about showing a project that you are deeply involved in to the mystical people behind the technology used, and being rewarded with many "oohs" and "aahs".

The second workshop I attended was hosted by Autodesk and titled "Creating Web 2.0 Applications on an Open Source Geospatial Platform". This workshop turned out to be a disaster from the start. The Autodesk folks distributed cd-roms to the crowd and requested that everyone in the audience participate in creating an open source web application from scratch. The CD-rom included installation files for OracleEX, Apache, Autodesk and other applications. To my dismay (or perhaps relief), the Autodesk team informed the 50 or so participants that this would only work in a PC environment, immediately making the 15 of us Mac users feel completely neglected. However, since I (with my Macbook), was smart enough to have Parallels running Windows XP, quickly got it booted up and proceeded to install OracleEX. After about 5 minutes of staring at a loading screen, I was then greeted by an incomprehensible error, of which the Autodesk rep gave me a grin and a shrug as if to say "you're a Mac user, I can't help you". Thinking that this was probably due to using Parallels, I then remembered that I was super smart to have installed BootCamp on my Mac, and therefore I would be able to run Windows natively. Quickly rebooting my Macbook into XP, I re-ran the installation file, and was again greeted by a different error message. This time it was telling me that I had run out of disk space. My bootcamp only had about 1.2GB of free space, and apparently OracleEX required 1.6GB of space.

So, after a thoroughly frustrating hour of trying to install software on my Mac, I was no better off from where I started, and we had yet to see an application demo. At the break, I decided to leave the room with other disgruntled attendee's. Outside, I met with 2 participants from Japan.

Satoshi Yamaguchi (left), works for Hitachi as their Researcher in Intelligent Media Systems Research Department. Tadayasu Sasada (middle), works for Mapion in their Corporate Planning Division. Satoshi took out his laptop and proceeded to show me a project that he and his team have spent the last 5 years developing, and it sure blew me away. It was a disaster management simulation model that ran on a Google Earth-like client that Hitachi has built from scratch. The navigatioin was entirely in 3D showing sharper topographic visuals than in GE. The simulation model he demonstrated was of a region in Japan that had recently suffered from a flooding disaster. Reminiscent to New Orleans, the model showed how he could zoom into an area close to a river, tilt the map, and click anywhere along the river to simulate the levies being breeched in that area. The model then displays a gush of water pouring out and flooding the area, tied with a time navigation represented by an analog clock. The algorithm behind the flooding is based on elevation data from Lydar imagery, which allows the application to calculate the amount and direction of water as it flows through the landscape. Satoshi is now on a "demo" tour, trying to market his application to the international community, hoping to secure some partnerships and funding to further his program.

Tadayasu, works for Mapion, a company similar to Yahoo, an ad-driven site which provides searchable mapping services (http://mapion.co.jp).

One of the projects that he has been working on is the creation of a light-weight, 3D web simulator.

http://labs.mapion.co.jp/koko3d/

The split screen interface shows a 2D map on the left with a red arrow showing current location and view direction. This is directly tied to the image on the right which shows a 3D rendering. Obviously, this is not a "true" 3D environment. However, the 3D image is produced dynamically from a server hosting some sort of 3D application that generates the images on the fly. Some smart caching allows the site to perform very fast on areas that have already been travelled upon. You can use the keyboard arrow keys to move the arrow on the left, and it automatically updates the 3D image on the right.

So, after my first day, I feel as though I have accomplished quite a lot. I have met with many folks from the Google Maps API team and made my pitch for the necessity of adding temporality to the API. I have also learned of some fantasic work done by my peeps in Japan. I look forward to what Day 2 has to offer.