I have generally been very happy with GIS in the Rockies (more on that in another post) as it’s good to get a look at different uses, concerns and perspectives. This afternoon, I sat through the “Collaboration Panel” discussion. The panel was made of of a few people representing state, regional and local governments and well as utilities and academia. Almost uniformly, there was a fear (yes, I mean fear) of crowdsourcing that was best summed up by the following statement:
“Crowdsourcing presents a vulnerability to us.”
Thank goodness Glenn Letham was sitting there and tweeted it or I might not have believed I heard it. Over the course of the discussion, talk kept returning to the image of “just anyone” editing spatial data, opening up potential liability for jurisdictions. Another gem:
“When everyone is responsible, no one is responsible.”
I will grant that the entire discussion did not focus on crowdsourcing but it was a topic that was circled back to several times. I was left with a sense that, when these people think of crowdsourcing, they literally picture crowds of the “great unwashed” making random edits to critical data sets from there iPhones. Fear, uncertainty and doubt was projected at such a level that it exposed, well, fear, uncertainty and doubt on the part of the panelists.
A great premium was placed on data being collected and managed by “authoritative” data providers, who collect data using a set of validated business rules and procedures. Yet a great premium was also placed on getting to the point where such data sets can be freely shared. This exposed to me another issue: a complete lack of understanding of the concept of crowdsourcing.
I realize that, for governments, there can be legal and statutory issues with publicly sharing data sets (the Privacy Act with regard to parcel data, for example) that must be worked through. For this discussion, I will set those aside. The following statement was made late in the session:
“OpenStreetMap will never be an authoritative data source.”
To that, I offer this (which I could not offer in the session since the floor was not opened):
If you feel you are the “authoritative” provider for street data for your region and that you capture that data using vetted procedures and business rules that assure data quality and accuracy *and* you profess to want to be able to share that data with the public, why don’t you donate that data to OpenStreetMap so that it would then be providing “authoritative” data (or at least data derived from an “authoritative” source). People who choose to use OSM for their applications would then be using high-quality data.
Why not, rather than tear down the concept, engage and participate, provide the model for how to do so and demonstrate leadership? In other words…
Be the crowd!
“why don’t you donate that data to OpenStreetMap so that it would then be providing ‘authoritative’ data”
We inquired on doing just this several times.
We rebuffed with assertions that our ‘authoritative’ data would hurt the mapping community and stifle community building efforts. We were told in no uncertain terms to take our data and go away. But hey, if we wanted to spend the staff time to edit in all 300,000+ street segments one at a time, we could do!
Hi Brett,
The U.S. Chapter of OSM is trying to change that an work with governments to use their data. I’m not sure who you talked to before. If you are still interested email me at kate (at) osmf.us.
My work here is done. 🙂
Sorry Kate, but I while I was the point person to try to get importing help, I do not make the decisions on data release. After the last attempt last month, the gatekeepers on our GIS data decided to focus on other projects.
I’m hoping to get a traceable WMS up some time, and I might be able to provide a copy of the data to someone directly with compatible licensing; but I don’t think I can get permission any more for the county to participate directly in an import.
Bill, interesting post, what a shame the floor was not opened for discussion.
May I suggest a different of your post: Lead the crowd!
Cheers,
Jw.
Jw,
Thanks. Just an FYI that your comment posted twice so I deleted one. I don’t know if it sends a message notifying of such things so I just wanted to let you know.
Best,
Bill
Bill, I wanted to write:
May I suggest a different end of your post: Lead the crowd!
Could not find that edit button on WordPress (like on LinkedIn)…
Cheers,
Jw.
Jw,
I certainly understand your sentiment but one view of crowdsourcing is that it is effective precisely because it isn’t led.
Of course, that’s also precisely why crowdsourcing’s detractors say it isn’t effective.
Bill
Bill,
You make some excellent points. However, I think a big part of the problem is that people wish to set aside the legal issues associated with combining “authoratitive” data with crowd-source data. In an ideal world it would be easy to combine the two types of data sets, but the reality is that there are significant legal issues (privacy, intellectual property rights, data quality/liability) that make it difficult to combine the two without serious thought and discussion. Until both (or all) sides are willing to reasonably articulate and address those issues I am concerned the potential benefits will not be realized.
Kevin,
I completely agree. I have spent a good part of the last nine years supporting data sharing efforts at the federal level and have encountered such issues first hand.
I think it can be useful set those issues aside in order to identify the goal but working through them is the hard work that is necessary to achieve that goal.
Where we have seen success, it involved a lot of serious discussion so I agree that is what is required. Unfortunately, I didn’t encounter that yesterday. I encountered attempts to propagate FUD, for which I have little tolerance.
It’s interesting to see the parallels with our TeamSurv project, using crowdsourcing for marine bathymetry data.
Putting aside technical discussions about accuracy, some are very enthusiastic about what we are doing, whilst others decry it as it is not official data. Yet the official data can often be over 100 years old – for example, the ast of Captain Cook’s surveys from the 1600’s were only removed from official charts a few years ago!
Is that where the disconnect is then? Maybe supporters of OSM need to do a better job of describing what it IS and what it is NOT. By not doing a good job of that, we may have the perverse effect of perpetuating the silos.
It wasn’t just OSM but the concept of crowdsourcing in general that caused heartburn. I found myself reconsidering “VGI” as a less-threatening term (to these people) to describe it.
I got over that quickly, though. 😉
All depends on authoritative at what level. In Haiti it is still serving as “the basemap”. The 1.3 million people that are displaced? They are being registered and that registration is geographically linked to objects in OSM. This is being done by the International Organization on Migration. Hard to argue that isn’t authoritative.
Looks at Europe as well. The most popular navigation application in Germany? Built on OpenStreetMap data.
You probably wouldn’t keep your parcel data in OSM. When I think of extremely accurate survey data I always think of parcels. We spend all this time getting really accurate property lines, probably most of time arguing where to put a fence. Less accurate data on the other hand often can save lives. It is all about what use you have for the data.
Interesting you should bring up parcel data, Kate. First off, I don’t think it has a place in OSM. Mainly because OSM is about what can be physically verified – on the ground – by anyone. Property lines are only nebulously ‘real’.
And thus far (as far as I know – someone correct me if I’m wrong) GIS hasn’t reached the point where it can be used to decide who owns what. This makes it rather poorly suited to delineate property lines or political borders.
In these parts, in fact, there’s still considerable confusion over ‘who owns what’ – the parcel delineated on the ground, or the parcel paid for according to what’s on paper? They almost never agree.
The point I’m trying to make (poorly) is that OSM isn’t about questions – it’s about answers. Concrete answers. A road is a road, and there’s no denying its physical presence. But political boundaries and property lines are outside the scope of OSM – they’re not our (the collective ‘our’) worry and not our problem. We’re concerned with the tangible, not the intangible.
Just let me walk down the road with my GPS on a sunny day, then upload that data to OSM. That’s the ‘Power to the People’ I’m talking about. The power to delineate the parts of the world that are meaningful to us.
I wasn’t suggesting parcel data in OSM either. I mostly agree, about the tangible. I’m not sure what the answer is with boundaries though, people expect them on a map. Knowing what town you are in can be important with some boundaries it is probably easy. For example the boundary between Virginia and Maryland is the Virginia bank of the Potomac, somewhat easy to delineate. I have no idea what happens when the river moves though.
Glad I wasn’t there, Bill. I might have done someone a grave disservice. Two points:
“crowds of the “great unwashed” making random edits to critical data sets”
– Absolutely true. And only a fool underestimates the power of the great unwashed (ask Marie Antoinette).
“People who choose to use OSM for their applications would then be using high-quality data.”
– Not ‘would be’, Bill. ‘Are’. I’ve written about this before.
Great post. Allow me to tell a blue bird about it, as well.
Thank you. Especially, thanks for the second point. Rereading that, it can sound as though I was implying that I think OSM’s data is not currently of high quality. Of course, I don’t think that but I appreciate the opportunity for clarification.
The same arguments that they have against wikipedia. Yet wikipedia hasn’t taken Britannica out of business, nor has it replace professors in universities or journals.
For many people used to having gatekeeper powers, it rattles them that some random person with can correct their work.
There will always be a place for specialized data, but OSM is going to become the base set for many future applications, just because it is free data. It low entry cost rewards the use of OSM in projects.
They actually mentioned Wikipedia. “How many times have you seen something in Wikipedia that you knew was wrong?”
To which I was thinking “Did you fix it while you were there?”