A Rebuttal to “Building a National GIS”
January 13th, 2009by Sean Gorman
A proposal recently came online recommending that the new Obama administration should invest over a billion US dollars to build a “National GIS”. The proposal claims “This system can be built quickly, immediately creating high tech jobs, and will serve as a public resource for project planners to support transportation infrastructure, water resource management, alternative energy research, and project siting.”
While I am a big fan of investment in geospatial technologies I believe we need to ask the question if this is the best use of resources and the right technology implementation. The claimed benefits are job creation, quick deployment, and utility as a public resource. So, we have a big price tag, over a billion dollars, so it can be assumed that whoever wins this contract with have to hire people to build and deploy the “National GIS”. The economic multiplier here is about one since you are building a specific system to deliver a National GIS to a single customer - the US government.
The second factor is “quick deployment”, and while I’m a bit skeptical on this point, I’m assuming this means there are GIS components that can be sold to the US government to build a “National GIS” quickly. The report goes on to say the “National GIS” would use “existing modern GIS server technology”, so I think it is safe to assume that ArcGIS server technology or other GIS server (is there another one - MapInfo maybe?) would fill this need. So, we have a proprietary technology driving a monolithic centralized system that delivers a large contract to a small number of vendors. After spending a billion dollars plus we have the hope it will deliver a public resource.
While this would be great if it did turn out, I can’t think of any current GIS projects that are a heavily used public resource. I don’t mean a resource for GIS users, but a resource for the public - the citizenry. Google Earth, Ushahidi, MapQuest, Local Live, Yahoo! Local etc. are all things the general public uses, but rarely GIS. So, who are we building a pubic resource for with our billion plus dollar?
I’d advocate a slightly different plan. Invest the money in making government data open and freely available to the public, when the data is appropriate, in popular open formats. Then let the community innovate around the data and create great applications with it. Let’s face it the government is just not a great builder of user facing Web technologies. When was the last time a government built a piece of Web technology that caught on like Twitter wildfire, ARPAnet? This is not meant as a slight just a question of allocating resources where you get the best return on investment.
Vivek Kundra’s “Apps for Democracy” is a great example of this principle in action. He made Washington DC data (including geospatial) openly available in standard formats then had a contest to develop applications around it. The program was a great success and received many accolades and Vivek has become one of the front runners for the new Federal CTO position.
A similar approach could work for building out critical components for the geospatial capabilities the nation needs. Provide grants as incentives to direct entrepreneurial companies to build prototypes to solve the outstanding problems. Then take the most successful pilots and deploy them to production with the funding and support to make them realities. Why put all your money in one massive project with a few vendors and hope it works out a billion dollars later. Just trust the lobbyists were right?
Most importantly, though, are the benefits of this type of approach - true economic stimulus and job creation. The majority of job creation comes from small and medium sized businesses. In this approach you are specifically incentivizing those companies to innovate to solve pressing problems resulting in multiple winners and dual use innovations. It is what America does best - hard driving entrepreneurial efforts to innovate and solve tough problems. If we want to get out of recession and stimulate the economy it is going to be by creating incentives to innovate and spark competition. I firmly believe it won’t be by lobbyist driven billion dollar projects to a small cadres of vendors/integrators. Pretty sure that is what got us into this mess.
Popularity: 14% [?]
January 14th, 2009 at 6:48 am
“Google Earth, Ushahidi, MapQuest, Local Live, Yahoo! Local etc. are all things the general public uses, but rarely GIS.”
Umm… How are you defining ‘GIS’ here? It seems to me that these applications are ALWAYS geographic information systems.
January 14th, 2009 at 7:42 am
Share your questions about wisdom of National GIS initiative: Virtual Alabama is THE example of how to do it, but building from local level up.
I’m co-authoring a book with Vivek on “Democratizing Data,” so I share your zeal for how that simple step can play a much more powerful role in fostering innovation & making public co-creators. This piece I wrote in TechPresident about the Obama stimulus package points out that he could really start the transparency ball rolling AND sell the stimulus by making disclosure a key part of the bill.
January 14th, 2009 at 9:11 am
Hi Terry -
Good point on semantics - I was defining GIS the way the authors of the proposal do. There have been several speeches where Dangermond and others have made a stark delineation between GIS and geo-content.
Maybe a better way to put it is none in the list (local search etc.) use “existing GIS server technology”. GIS technical folks use “GIS server technology” not the general public. I’m sure ESRI would like to change that but it is certainly not the status quo. Again the problem with a one size fits all solution. I’d say use the best technology for the task at hand. For high end GIS analytics ESRI is probably your best bet, but for public consumption of data and analysis not so much (Google/MSVE etc. have been far more successful).
@David - look forward to checking out the book sounds great. Virtual Alabama is another great example of breaking out of homogenized solutions. There seems to be a lot of ground swell around transparency and hope it continues post inauguration. You might be interested in:
http://barcamp.org/Government20Camp
At one point I think it was even called Transparency Camp.
January 14th, 2009 at 10:05 am
I get a little more disgusted by this proposal the more I read it. It’s sort of a skeevy come-on: “I know you’re having a lot of trouble with your money situation, so I’m here to sell you a product that’s going to fix your life up straightaway. Just send $1.2 billion and a self-addressed, stamped envelope to Jack @ Redlands, CA.”
I’m with you, though, in that I’m very disappointed there’s nothing in this document about access specifically. The thesis is that an investment of this kind will result in framework layers, data update procedures, data management, “tailored applications,” new bureaucracy to oversee it all, and “the necessary technology to support” it all. I guess it just slipped their minds that the ultimate, primary benefit is that the country’s unbelievably innovative (and vast) stable of developers and scientists would be able to access this stuff in ways that make sense (APIs, standards-adherent protocols, etc.). Maybe they assume it’s implicit, but to me it betrays the true interest here: get our stuff sold to a desperate, reeling new government.
January 14th, 2009 at 10:33 am
It should be completely possible to aggregate, store, and serve up national GIS data and applications using OpenSource software.
If the true need is to collect and develop a national data set, put the money into paying GIS professionals and a group of AmeriCorps, GISCorps, etc. to pull the data together.
There is no need to send a lot of money into the pockets of a privately held company that sells expensive, proprietary GIS software that serves up data in proprietary formats.
If there are OpenSource GIS projects that show promise, but are not up to ‘best of breed’ yet, why not fund development of those projects. The money goes directly to individual workers, the investment is leveraged because the resultant product is available for free. This allows consultants and contractors (small/medium business) to take free product and create applications for other businesses.
Why should the American people have to pay twice for their data or GIS software…
January 14th, 2009 at 11:55 am
Well put Sean! Should Jack and his cadre win this, it will not only give them a huge project to ride out this recession (fine, I can respect that), but will also line them up for every future job extending what will become some “National GeoArchitecture” that will be (by design) complex enough to keep other vendors at a severe disadvantage.
I think that a hybrid solution, based around an spatially enabled database system (Oracle Spatial, SQL Spatial, PostGIS depending on the scale) would be a better model. Spread that billion around and setup standardized ETL systems that can aggregate from local to state to federal systems across DBMS platforms. That alone would be a huge improvement.
While a system like this may not have all the “optimizations” (read: lock-in) of a single-vendor COTS solution, parts of the system could be replicated out at minimal cost (i.e. get the data for a State, in say PostGIS format, with pre-authored SLD’s for use in GeoServer). ArcGIS Server should still be in the mix, but should be used for it’s strengths - the ability to do complex geospatial operations. But cached and dynamic map layers can be served via open source software that has a nominal cost to scale out.
With the COTS cost savings from using a hybrid architecture, money could be allocated to funding start-ups, which would be selected via a Vivek’s model - “Here’s the data, now come up with innovative ideas on how to use it, and we’ll fund your start-up”.
Make this be truly “free” money. How many entrepreneurs would jump at the chance to keep VC’s out of the first round, and keep ownership to themselves? That incentive would drive innovation, and that’s a big part of pulling us out of this economic dip, and has always been a big part of America’s success.
Cheers,
Dave
January 14th, 2009 at 1:57 pm
Great points and I think opensource and open standards would have a huge roll in a successful program. If it is an open competition we know clever developers are going to leverage opensource to create solid solutions.
The beauty of the advancement the community has made with open standards and API’s etc. is that you don’t need a single vendor in order to achieve interoperability and optimization. ESRI can still be a part of the equation - just not the whole equation. Let them excel where they have the best solution.
Great point on using grants and pilot projects to provide capital to start ups and SMB’s. The VC community is hurting and raising capital is tough, not to mention would be great to hold off on a round or replace it all together.
January 14th, 2009 at 4:59 pm
[…] Gorman has a good rebuttal over here. And I poached the following quote by Sean from James’ original post… I don’t think […]
January 14th, 2009 at 11:52 pm
I read that infamous pdf with a so-called “proposal”. It took me 30 seconds.
So now saving the mapping industry with a prehistoric of a “project” will save the american economy.
Come on guys, you can do better!
January 15th, 2009 at 12:12 am
Is ESRI going to bankrupcy that it so desperately needs a lifesaver,, or is this a last attempt to make GIS useful not by serving to planning purposes and the like but by supposedly being the core of an industry whose destiny was to save the economy?
Quick deployment with ESRI technologies… you gotta be kidding
Ever heard of open and free geodata? Ever heard of SDIs? ever heard of Open Source GIS?
January 15th, 2009 at 5:01 pm
@SeanGorman
http://gov20camp.eventbrite.com
and
http://transparencycamp.eventbrite.com
are two different events…
January 16th, 2009 at 11:07 am
I am so glad to see the backlash on this and from the very ppl that can actually get the job done. As my cover indicates, we are forced to work through the so-called bandits because they hold the procurement strings. They mark up our work and provide little value for that margin - other than having greased the financial skids. And in many cases stand in the way of our ability to get the gov’t the solutions they really need.
And isn’t it time we had a new icon for our industry - Jack brings out the ire in most veterans of the biz that I know.
We are entering “the era of change”. We probably can’t do much to fix the financial crisis but I think we can all at least try to break the status quo in our own industry and fight for open, interoperable, and low cost geospatial tech. And my own personal peeve - stop the talking (policy) and start the walking (develop content/products/tools/solutions).
January 16th, 2009 at 12:37 pm
I think opensource is probably the best return for the investment of public dollars instead of going to a proprietary system with a good product but not always the best.
I have used several excellent tools such as MapServer, MapWindow, GRASS, Quantum GIS, UDig. If the government would make smaller grants to such efforts, it may bear more return for the dollar.
Maaza
January 16th, 2009 at 1:12 pm
It Sounds like you are proposing an incentive program very similar to Canada’s GeoConnections program:
http://www.geoconnections.org/en/index.html
January 18th, 2009 at 12:15 pm
[…] A Rebuttal to “Building a National GIS”‘ by Sean Gorman, co-founder of Fortius One - the folks that brought you the GeoCommons and (Map) Finder/Maker n- is available at: http://blog.fortiusone.com/2009/01/13/a-rebuttal-to-building-a-national-gis/ […]
January 19th, 2009 at 2:19 pm
[…] First Mr Dangermond writes a pseudo-proposal to the new government. It will, no doubt, save America´s economy. Apparently, all it takes is to create a National GIS! (and ESRI technologies are just […]
January 21st, 2009 at 5:45 pm
Thanks Peter - I’d heard a rumor one had been replaced with the other great to hear they are both still on. See you at TwinTech tomorrow - looking forward to it.
January 23rd, 2009 at 7:11 pm
While there are many great alternatives proposed here along technology and business lines I’m afraid they are missing a critical element of the GIS.
First, Like it or not, Microsoft, Yahoo, MapQuest, Google and even ESRI use mostly commercial data to make their applications work. Public sources of data vary widely in quality and reliability. This impacts their usability in applications. We forget government collect and create data to satisfy their own business needs making it useful for some applications but useless in others.
We should recognize the work of openstreetmap, I suspect the majority of the money be spent on the development of critical datasets with much funding going to develop data for GIS have-nots
January 25th, 2009 at 3:35 pm
January 23, 2009 - Today a collaborative grassroots coalition released a concept paper titled “NSDI 2.0: Powering our National Economy, Renewing our Infrastructure, Protecting our Environment.”
The NSDI 2.0 Concept Paper is available at http://www.nsdi2.net.
The paper notes that “the American Recovery and Reinvestment Plan will create jobs in the short-term and spur economic growth and competitiveness in the long-term. But this plan must be designed in a new way. We must make smart strategic investments that serve as a down payment on our long-term economic future, create millions of new jobs - and provide the American workforce with new skills.
To build a 21st century economy, we must engage Local, State, and Federal agencies and their partnering contractors across the nation to create jobs rebuilding crumbling roads, bridges, electric grids and schools - but these organizations need an updated online information network that will allow them to rebuild in a smart, efficient, environmentally conscientious and sustainable way. A National Spatial Data Infrastructure (NSDI), updated with vital environmental information, will speed economic recovery by producing jobs putting “green” shovels in the ground quickly and tie together ongoing government initiatives.
An “NSDI 2.0″ will leave the country with a public resource, a modern spatial data infrastructure that will become a foundation for new business and technology investment - including broadband infrastructure development efforts now under consideration. Most importantly, this framework provides a sustainable, long-term infrastructure and innovation investment that will create thousands of new jobs and contribute to the economy for many years to come.”
To learn more visit: http://www.nsdi2.net.
About the NSDI 2.0 Concept Paper
The NSDI 2.0 Concept Paper is a proposition offered by a collaborative grassroots coalition that advances a business case applicable to the entire geospatial and environmental sector. It embodies an inclusive, collective approach that is well positioned to provide widespread economic success throughout public, private, and non-profit organizations of all types and sizes. This paper represents (although unofficially) the consensus view of a collaboration between corporate entities, the non-profit sector, the open source GIS software community, as well as Municipal, County, Regional, and State Government Agencies who produce and rely upon the Nation’s critical geospatial and environmental data resources. For more information please visit: http://www.nsdi2.net.
February 4th, 2009 at 12:59 pm
I’m seeing a lot of Neo-geographer reactions here (not meant as an insult). As one who bridges both worlds, I can see a great deal of good in a National GIS. No, ESRI’s Arc isn’t necessary and running the whole thing off a DB backend would be great (if slow).
I’ve worked in dozens of states, with dozens of local survey companies. I’ve worked with Universities around the country. As one commenter already said, “Public sources of data vary widely in quality and reliability.” That is putting it lightly! Surveyors must be licensed to get their data legally accepted. It is very good, but getting at it is impossible. For all of us who have to go state-by-state and more often county-by-county, a good, National GIS would be a boon to our industries. It would be HUGE. I, for one, would love to see one.
January 18th, 2010 at 5:11 pm
[…] A Rebuttal to “Building a National GIS” (Sean Gorman, January 13, 2009) […]