Thursday, July 30, 2009

My GeoWeb 2009 presentation

Here's a video of my GeoWeb presentation on "Building a new location-aware infrastructure for calendaring and scheduling". Recorded and edited using ScreenFlow.

GeoWeb 2009 presentation by Peter Batty from Peter Batty on Vimeo.

I'll post more thoughts on GeoWeb soon - in the mean time you can check out my tweets and the geoweb stream on twitter. And there are some pics of the social event here.

Tuesday, July 28, 2009

Are we now in the post-neogeography era?

So all the discussion about my last post got me to thinking about how to define "neogeography". I was really using it in a pretty broad sense, as a convenient label for the “new generation” of web mapping tools and related technologies that have emerged over the past few years – essentially the “disruptive technology” that Clayton Christensen talks about. These systems include Google Maps and Earth, Virtual Earth, Yahoo Maps, and various offerings from smaller players. However, I feel that the distinction between “neogeography” and “GIS” is becoming increasingly blurred, to the point where it really doesn’t make much sense to even try to make a distinction in many cases.

As I discussed, “neogeography” systems are now increasingly being used for more activities that were in the domain of “traditional GIS”, including cartography, analysis and data creation. And “GIS” is increasingly using ideas that originated in neogeography systems, like using pre-rendered raster tiles for fast map rendering, implementing REST APIs, etc. The summary for O'Reilly's "Introduction to Neogeography", which is probably as good a source as any, says "Neogeography combines the complex techniques of cartography and GIS and places them within reach of users and developers". So according to that definition it's really about simplification, which is certainly a key characteristic of systems like Google Maps and Virtual Earth. But it's obviously not a hard and fast definition, and many efforts within the traditional GIS space could be said to have done the same thing, especially more recently as ideas have been taken from the first round of "neogeo" products.

I think that one other difference between those who describe themselves as neogeographers and those who describe themselves as GIS professionals is that the former see geospatial data as just another data type, whereas the latter tend to see it as something more specialized. But this also a continuum, and this difference in philosophy has been around in the industry to a large extent since well before neogeography arrived on the scene (but maybe with a smaller proportion of those with the first opinion).

So to me, the main point is that there is now a broader set of tools available to help address geospatial problems, regardless of what you call them. I think that many people in the traditional GIS world haven’t had too much exposure to these new tools, and a lot of people still have the mindset that they have selected a primary GIS vendor – ESRI, Intergraph, Smallworld or whoever – and they should only use geospatial software from that vendor as it will make their life easier. But it is really very easy to mix and match geo software tools these days, so it’s worth looking around at the various new options. But clearly there are still many tasks for which the best choice remains the current generation of traditional GIS. However, the traditional GIS vendors will face major price pressure over the next few years as the newer generation of simpler and cheaper systems mature.

So in summary I really don’t see much value in trying to define some tools as “neogeography” and some as “GIS” – to me it’s all just geospatial technology and you should choose the right tools for a given job. Maybe the term "neogeography" has outlived its usefulness and it should just quietly fade away, except as a historical term to describe the new geospatial systems that emerged in the past few years, before they interbred with traditional GIS and became just part of the new and richer set of tools that is now available for geospatial applications? Or we just regard terms like neogeography, GIS and geospatial technology as synonymous?

Twitter discussion on "neogeography" versus "GIS"

My post yesterday triggered some interesting discussion on Twitter (I guess the cool kids don't comment on blogs any more, that's so 2008!). Here's a quick cut and paste with most of the relevant stuff (the relevance fades towards the end but I figured I would leave some of those bits in!). I will post more thoughts shortly on what I think "neogeography" actually is. If you want to figure out who someone is on twitter, just go to twitter.com/username - for example I'm at twitter.com/pmbatty.

cageyjames: @pmbatty isn't all that stuff still very niche? I'm just not seeing the quick pickup you are I guess.
cageyjames: @pmbatty That said I'm coming off #esriuc and you #sotm so we skew on our last experiences. Reality is probably between us.
cageyjames: Of course Neo and GIS are just labels people put on others to limit their growth.
pmbatty: @cageyjames early days still but my point is more rapid growth of neo in areas like cartography, analysis, etc than many people think
pmbatty: @cageyjames I agree I'm not a big fan of labels like neo, paleo, GIS, it's all the same problem space
cageyjames: @pmbatty but can you call that neo? What Stamen is doing isn't neo in my book. You seem to say innovative is neo. That isn't true at all.
cageyjames: @pmbatty When I was taking carto classes back in college, was I neo because I used Freehand over pen and paper? Not in my book.
pmbatty: @cageyjames Stamen is using OSM, Mapnik, Virtual Earth, etc - would think that was "neo" though we agree the label is a bit pointless
geobabbler: @pmbatty Neo is GIS. Just like "desktop mapping" was.
pmbatty: @geobabbler I agree that Neo is GIS - I think that is what I was trying to say but you were more concise :)
geokaren: @pmbatty @geobabbler I think of neo as less rooted in technology and more the concept of moving it beyond traditional GIS
cageyjames: @pmbatty ESRI uses OSM, Virtual Earth and other "neo" packages. That doesn't make them neo. Tools make not neo.
geobabbler: Neogeography represents the intersection of GIS with the mainstream web. It's the web telling GIS "this is how you do it."
geobabbler: @pmbatty and I was trying to agree with you but not very well. :-)
pmbatty: @cageyjames Not clear on your definition of "neo"?
pmbatty: @geobabbler Good, I think we're agreeing we agree - 140 chars is a limitation sometimes :)
SeanGorman: @cageyjames @pmbatty 1) getting GIS stuff to work in a browser 2) making GIS stuff accessible to non-GIS people = current innovation
pbissett: @geobabbler @pmbatty I sometimes wonder if we shouldn't maintain an Geo-IRC channel for such discussions. Is there interest?
SeanGorman: Tough bit will be doing new things GIS or others have not thought of - handling/analyzing/visualizing big data and mobile data = my bet
geobabbler: @SeanGorman The "new things" will arise from the questions asked by new audiences as GIS continues to diffuse.
cageyjames: @pmbatty neo is a label, nothing more
SeanGorman: Serving tiles for viewing imagery/maps seems the biggest innovation that GIS was not doing at all followed by user generated geo-content
cageyjames: @SeanGorman it is all a moving target isn't it. What was once hacker is now helping me locate coffee shops near crime.
geobabbler: @cageyjames spot on.
geobabbler: @pbissett it has been bandied about before.
geobabbler: it's somewhat ironic that I am engaged in a "neogeography" discussion while waiting for @ajturner.
pbissett: @geobabbler If I remember correctly from "Innovators..", it was the search for new solutions to old problems that caused the disruption.
SeanGorman: @geobabbler tell him has to wear a wig every time he is referred to as the father of neogeography
SeanGorman: @geobabbler possibly some knickers and a waist coat as well
SeanGorman: @cageyjames yes - because I want to be perky when I get mugged. If life was stationary it would be quite dull.
pmbatty: @geobabbler @SeanGorman I was just going to say that @ajturner will explain all this to you, he is our father :)
cageyjames: @SeanGorman classic, now give me the GeoRSS feed!
SeanGorman: @pmbatty hahahaha I think that quote now makes him Darth Vader or possibly Darth Helmet
SeanGorman: Ok I'm going to quit heckling before @ajturner brings up the fact I can't write code and that I'm of dubious use in general
jharpster: @SeanGorman I bet if we brought out some VBA code you'd be all over it. Don't take @ajturner too seriously.
jharpster: Blog by @pmbatty on the influence of neogeography. "people in the traditional GIS space remain largely unaware" http://bit.ly/21QiKC
ajturner: consider this corollary for those asking what 'neogeography' is: is using an iPhone to update Facebook the same as 'Information Technology'?
cageyjames: @ajturner yup, my point exactly. We are all neo and all paleo. Thus labels fail us when applying at a high level.
fantomplanet: @cageyjames I once heard @ajturner used walk to school uphill both ways!
fantomplanet: @ajturner Answer to your corollary: It's an element and activity that is part of secondary orality (http://bit.ly/10SJeQ). Same with carto.
fantomplanet: @SeanGorman Yeah, but I can still code... A little. #fullspectrumgeographers
SeanGorman: @fantomplanet bragger - I programed my VCR last week to tape "the View" while I was at work.
cageyjames: @fantomplanet coding is for sissies, you get hair on your back by managing projects.
pmbatty: Finished beer and interesting conversation with @perryevans, solved world's neogeography problems, now time for dinner :)
fantomplanet: @pmbatty "Solved world's neogeography problems?" #neogeodeathakawherecamps?
pmbatty: @fantomplanet I just hope I can still remember all the solutions tomorrow :)
geomantic: @pmbatty @SeanGorman @geobabbler @cageyjames I'm late to the neo paleo geo GIS discussion, but see you found agreement 140 chars at a time?
pmbatty: Wish I was clever enough to understand @ajturner, the Yoda of neogeography :) (@SeanGorman: gift suggestion http://tinyurl.com/l6z62g)
fantomplanet: @pmbatty All @SeanGorman needs to buy @ajturner is a burlap sack and turn it in-side-out. #neogeoyoda
geomantic: @fantomplanet @pmbatty @SeanGorman ...or buy @ajturner a monk's habit. Oh, and get him an appt for a Friar Tuck haircut, too. #picturethis
fantomplanet: @geomantic I think @ajturner already has the haircut.
geobabbler: @ajturner it looks like @gezjames found your wig.
DruidSmith: With all this #neogeography kum-ba-ya, I'm surprised we haven't yet seen any choice words from @fakeajturner entering the fray...
fakeajturner: What? Have I been summoned? I was too busy showing the neogeo "who's your daddy"

Monday, July 27, 2009

How "neogeography" is rapidly moving into the "GIS" space

I thought I would do a post or two to set some context for the upcoming geothought “top 10 most influential people for the next 5 years” list, that I recently posted about. In summary, I see the next 5 years being a period of major disruption for the traditional geospatial industry.

As I have mentioned many times over the past several years, Clayton Christensen discusses in his book The Innovator's Dilemma how "disruptive technologies" start out focused on a different market, and don't address all the needs of the current "mainstream market" (traditional GIS in our case), but over time their capabilities increase to the point where they can address the needs of the mainstream market, typically in a way which is much cheaper and simpler than the previous technology. This disruption is now happening very quickly in the "neogeography" space, even though many people in the traditional GIS space remain largely unaware of it.

At several conferences I have attended recently - Where 2.0, WhereCamp and State of the Map (SOTM) - I have been struck by the amount of activity and innovation in areas that would have previously been regarded as firmly in the domain of "traditional GIS". I'll mention three: cartography, data creation and analysis.

There were multiple workshops and presentations on cartography and how to design nice looking maps at Where 2.0 and SOTM. For example, look at this presentation by Andy Allan on "Advanced OpenStreetMap Cartography":

Not only are the maps shown in this presentation generated using (free) tools like Mapnik that probably most traditional GIS people have never heard of, they are all using crowdsourced data that was captured by "amateurs" for free. Still think you need to be a highly trained GIS professional using sophisticated and expensive software to create professional looking maps?

Stamen Design is a company that presented at both conferences that has done some really impressive work creating beautiful maps and web sites - for example check out their Oakland Crimespotting site (which also uses OpenStreetMap data) and this beautifully designed hurricane map for MSNBC, which I think does a fantastic job of making it really easy and intuitive to dynamically filter different aspects of the data. Stamen has also created the innovative Walking Papers site, which is a brilliantly simple way of allowing people to update OpenStreetMap using paper maps - see a video and slides (PDF) of Michal Migurski's presentation at SOTM (which also briefly covers some other interesting projects). I have seen similar ideas to Walking Papers in the traditional GIS world, but as with many of these neogeography projects, what stands out with this is its simplicity and ease of use.

Another great example of progress on the data creation front is of course OpenStreetMap itself. I really like the browser-based Potlatch editor which is currently the main browser-based tool for editing OSM data. It takes a little learning (though not much by traditional GIS standards), but combines fairly simple functionality on the surface with some pretty sophisticated capabilities underneath which dynamically create network connectivity as you create objects, with easy ways to split and merge roads, etc. This is all functionality that until recently has required a sophisticated GIS - now all you need is a web browser. The next step forward in this area should be the Mapzen editor from Cloudmade, which if it lives up to the promise of its screenshots and description should be a huge step forward in usability for data creation and editing tools.

On the analysis front, many people are doing interesting things, one of the most notable being FortiusOne with their GeoCommons site and GeoIQ product. I am a big supporter of their philosophy that easy to use geospatial analysis tools should be available to anyone, whereas there is still a prevailing view in the traditional GIS industry that this should be the domain of specialized "GIS professionals". Several of the Stamen projects, and a variety of other presentations at Where and SOTM, included interesting geospatial analysis without the use of traditional GIS.

I could go on with more examples but this gives a general idea of how many new and innovative players are coming into the application space previously occupied by traditional GIS vendors. These new systems are in most cases some if not all of the following:
  • Easier to use
  • Easier to implement
  • More scalable
  • Free or cheap (often open source or with an open source component)
So big change is coming, and my list of the top 10 influential people for the next 5 years will be focused on those who are making this happen! To traditional GIS people who still think that neogeography is something different from what you do, I really encourage you to look more closely at some of these new tools and companies. I would also say to neogeography people who are working in these areas, try to spend some time chatting to a friendly "paleo" or two - while there is value in taking a fresh approach to these problems, there's also a lot of relevant experience that has been accumulated over the years that may be useful and avoid some reinvention of wheels (which I've seen a little bit of - though the benefits of a fresh approach far outweigh the drawbacks).

Friday, July 24, 2009

Silly poll of the week

So this week Directions magazine published the results of their poll on who will be the "most influential people in geospatial" over the next 5 years. I like Joe and Adena and they do a lot of good articles, but I thought this one was a bit silly, and even if you accept the silliness had some glaring omissions (and not just myself ;) !). Though there were some interesting nuggets in the published comments. However, if you treat it just as a bit of fun, it's a good way of generating some interesting conversation I guess :) !!

The question itself though is a bit like asking who will win the Superbowl in 5 years - who knows? If you can answer either question I suggest you head to Las Vegas and make a quick fortune on the betting tables. If you look at the past five years, I would say that unquestionably the most influential organization in the geospatial industry has been Google (it's hard to pick a single person from there). Five years ago they weren't even in the industry - they hadn't yet acquired Keyhole and hadn't launched Google Maps. But the whole landscape of the industry has completely changed since their arrival. So if you asked the same question five years ago, you would have entirely missed the most significant influencer. Now I guess that doesn't mean it's not fun to speculate of course, but a poll (as Joe somewhat acknowledges) is not the best way to pick up the more interesting up and coming people, who by definition many people aren't aware of yet. If you are going to do a poll like this, a more realistic question is who are the most influential people today, or over the coming year - which in reality is probably what this poll boils down to anyway.

In the glaring omissions category, the most obvious individual to me is Andrew Turner; I don't see how you can not have Microsoft in the list in addition to Google; and when looking at Google I would think you have to include Michael Jones - though again it's hard and somewhat arbitrary to pick a person or a small number of people from there. As someone commented in response to the poll, there is also a strong case for Vivek Kundra and/or Barack Obama too, in terms of what they are doing to make government data more easily accessible in the US.

Anyway, this all got me to thinking who I would put in my top 10 people likely to be influential over the next 5 years, and so I have decided that next week I will publish the "geothought top 10 people to watch in the geospatial industry" list. I have a pretty good idea of most of my choices, but if you have suggestions or want to send bribes etc, please let me know :) ! Currently I think that 3 of the Directions top 10 are likely to make the list, and the rest I think were not in the 22 nominees. But we'll see, it will be a bit of fun :).

Wednesday, July 22, 2009

Freezing whereyougonnabe and moving on to new things

I have been trying to raise investment funds for Spatial Networking for some time now but despite some interest still haven’t closed anything, unfortunately – it’s a tough time to be raising money. After weighing up all my options, I have decided to stop new work on whereyougonnabe (for now at least) and move on to some new things. I continue to believe that we have a good idea and have developed some valuable technology, and that there will be a strong market for future location and location aware calendars at some point – and it is quite possible that the work we have done in this area will be revived somehow.

However, one of the factors in my decision is that there are also a lot of other really interesting things going on in the geospatial industry right now that I would like to get more involved in, and it’s hard to do everything :) !! I have been working on several things in parallel with whereyougonnabe over the past year or two, including my role as Chief Technical Advisor at Enspiria Solutions, being on the Advisory Board of PublicEarth (both of which are ongoing activities), and working with Netezza on the launch of their spatial capabilities last year. I also did some interesting work with the Bill and Melinda Gates Foundation last year, to help start what is now the AgCommons project.

Some of the main areas that interest me right now include:
  • The disruptive transition from the previous generation of GIS to the “neogeography” systems, which is well and truly under way – I will blog more about this shortly
  • The growth in free and open source geospatial software
  • The huge potential of crowdsourced data, most notably OpenStreetMap
  • In a somewhat different direction, the Smart Grid, which is something I have been working on at Enspiria and which will be bringing major changes to the electricity industry, an industry I have worked in a lot during my career
So I’m on the lookout for interesting opportunities in these areas (or others). I am quite interested in additional advisory or part time roles that I could pursue alongside the existing part time work I’m doing, which would give me some flexibility to dabble with some other interesting technology ideas on the side. But I’m also open to considering the right full time role. If you'd like to discuss possibilities please get in touch!

I’d like to thank my fellow developers Glen Marchesani and Nate Irwin for all the effort they have put into whereyougonnabe – and as I said, even if it goes dormant for a while, it may re-emerge in some form in the future! We are still weighing up whether to keep the current whereyougonnabe system up and running for a while, or take it down, and will notify users what we decide within the next few weeks.

I’ll be talking at GeoWeb next week about what we’ve done with whereyougonnabe, why I continue to see it as an interesting space, and some lessons learned from the whole experience.

Thursday, July 16, 2009

OffMaps for iPhone review

On my recent trip to Amsterdam for the OpenStreetMap State of the Map conference, I made extensive use of the OffMaps mapping application for the iPhone - which incidentally uses OpenStreetMap data. The big advantage of OffMaps compared to the standard Google Maps application is that it can run offline, which is really important if you are going abroad, as data usage charges make it prohibitively expensive to use an online mapping application (you quickly get into hundreds of dollars worth of data charges!). The functionality is fairly simple - currently it's basically map display, without routing or search for points of interest (though these things are planned for the future, I heard from Felix Lamouroux, the main developer, who I met at the conference). But it makes good use of the GPS and the compass in the iPhone 3GS, which make it very intuitive and easy to use when walking around. Performance is very good since all the data is local. There is a simple interface for choosing the data to download and store locally (which you want to do ahead of time when you're at home, or connected via WiFi). You can run it in an online mode too, where it will dynamically fetch data over the network if it isn't already stored locally.

The OpenStreetMap maps for Amsterdam were very good quality and nice looking. This application illustrates one of the key advantages of OpenStreetMap over Google Maps and other commercial solutions, which is that Google licensing prohibits you from using the data offline. This fact, plus the availability of the Cloudmade iPhone Maps Library, is making OpenStreetMap data a popular option for mapping applications on the iPhone - just today, Dopplr announced a new iPhone application which also uses OpenStreetMap data (though strangely it uses different rendering from the other applications I've tried, and none of the freeways in Denver show up, though they do in the other applications - but I'm sure that will be fixed shortly).

So I would strongly recommend OffMaps, especially if you're going on an overseas trip - I am currently downloading data for Vancouver in preparation for my upcoming trip to GeoWeb, so if you're there and want to check it out please ask me and I'll show you!

Wednesday, July 15, 2009

WhereCamp5280 - cool event coming up in Denver!

We’ll be holding a fun, cool, educational and FREE geospatial event in Denver on August 14-15 – WhereCamp5280 (for those not from round here, 5280 is the elevation of Denver above sea level in feet – exactly one mile high!). If you’re already familiar with WhereCamp, you probably don’t need any more persuading to come along. The event is being organized by Eric Wolf, me, Ben Tuttle, and anyone else who would like to help out!

For those not familiar with WhereCamp, it’s an “unconference” on the subject of location, mapping and things "geo". An unconference is basically an informal conference which is organized on the fly by the participants. I’ve attended two unconferences now, both of which were excellent. The first was held by FRUGOS (Front Range Users of Geospatial Open Source), which you could also consider a predecessor to this event – though we intend for this to cover not just open source solutions (though that’s up to the participants, of course!). Here are a couple of reports on the FRUGOS unconference by me and Charlie Savage. The second was WhereCamp in San Jose this year, which I found really interesting and fun. I have to say that based on these two unconferences, the quality of presentations has been at least as good if not better than the “formal” conferences I have been to.

We have already had expressions of interest in attending from a wide variety of interesting geo-people, both locally and from out of town, so I’m sure there will be lots of great discussions. For an up to date list of people likely to attend check here.

If you are potentially interested in participating, please sign up to our Google Group and add your name to the likely attendee page. We encourage you to list topics that you would be interested in talking about or leading a discussion on, but this is definitely optional! While the agenda will certainly be flexible in unconference style, we plan to get some ideas together in advance to help us make the most of our time during the two days.

If you’re interested in helping out, please let me or Eric know. While DU has kindly donated the rooms for the event, we are aiming to get some modest sponsorship to be able to cover other costs including:
  • A fun social event on Friday evening (probably either at the Wynkoop Brewing Company, or my loft which is right above the Wynkoop)
  • Cost of projector rental from DU (which is not free) – if you have access to a projector(s) that we could use on Friday and/or Saturday, please let me know
  • Lunch on Friday and Saturday
Please see the Google Group for more info – your questions, suggestions and assistance are appreciated! We really hope to make this a good venue for exchanging ideas between the “neogeographers”, open source users, and users of traditional GIS (and some of us, like me, fit into all of those camps). If you work in the geo area in Denver, tell your manager that this is the best free education you can get on the latest developments in the space!

Tuesday, July 14, 2009

My presentation on "Geodata creation: past, present and future" at State of the Map 2009 #sotm09

A video of my presentation at the recent OpenStreetMap State of the Map conference is now online at Vimeo. There's also a copy of the slides at SlideShare. I talk about the four major business models that have been tried in regard to creating geodata, and how they are all handicapped by the very high costs involved when you use the traditional approach which involves paying the people who create the data. And I discuss how crowdsourcing completely changes things by reducing this labor cost to zero. From everything I saw at State of the Map, I am convinced that use of crowdsourcing in general, and OpenStreetMap in particular, is going to massively grow over the next couple of years.

Geodata creation: past , present and future - Petter Batty (Spatial Networking) from State of the Map 2009 on Vimeo.

Other videos are being posted online at the sotm09 channel on Vimeo - I recommend you take a look. There are lots of great presentations, but one that I guarantee will make you smile is this 5 minute one on Mapping of Historical Sites in Japan - check it out :).

Monday, July 13, 2009

Quick report on OpenStreetMap State of the Map 2009 conference

This is just a quick initial report on the OpenStreetMap State of the Map conference, which has just finished in Amsterdam. First I'd say read Steven Feldman's summary, which I completely agree with. As I said in my keynote talk here (see the slides, video should be online at some point soon, and I'll summarize it in a future post), the industry has always been hampered by the very large cost of creating and maintaining geodata. Despite all the technological advances over the years, this remains a very labor-intensive process and therefore is just fundamentally expensive by traditional methods (where you pay the people doing the data capture).

Crowdsourcing changes the paradigm by having volunteers contributing their time, and having access to free data (without complex licensing constraints) completely changes the economics of developing geospatial applications. Obviously the first reaction of most traditional geospatial people is to ask whether you can get good enough data quality using this approach. Dr Muki Haklay presented a very thorough and rigorous analysis of the quality of OpenStreetMap data against Ordnance Survey data in the UK, and his conclusion was "OpenStreetMap quality is beyond good enough, it is a product that can be used for a wide range of activities" (not in general - today - for very large scale mapping, but for small or medium scale mapping - the type of applications that today might use today data from Google or Microsoft, NAVTEQ or Tele Atlas).

The conference itself had a tremendous buzz about it, with 250 people from thirty-something countries (I think), and great enthusiasm and excitement from everyone participating. The presentations were a great mixture of people talking about really cool and innovative technical things, and heart-warming and moving stories about people creating maps from all parts of the world - in many parts of the developing world, OpenStreetMap is way ahead of any other data source.

Anyway, I'm off to be a tourist in Amsterdam for the day, I will write more in due course. If you are attending the ESRI user conference this week I really recommend you seek out the OpenStreetMap booth to find out more about what they are up to. And thanks to everyone who attended State of the Map for making it such a fun and inspiring event!