Everyone loves a laser
It’s been really interested to watch how the recently published Environment Agency (EA) LIDAR data has been seized on by a variety of communities to create interesting, fun and useful tools.
For many people, myself included, learning about LIDAR has been an interesting experience. It’s a technology that suddenly seems to be everywhere and which can be used for a number of different purposes, including helping cars see the world. For a geek perspective it’s a classic example of an exciting open data release. It’s not just that the data wasn’t previously accessible. For many people it wasn’t obvious that this type of data even existed. Happily, as it turns out there are plenty of tools with which we can use it. Cue lots of interested hackers playing with laser data. Everyone loves a laser, especially the green ones.
I think it’s going to be difficult to top that excitement with further releases. But that’s fine, because open data isn’t about finding the next most exciting new data. Its about publishing what we have. Sharing what we know. And then using that information to solve problems.
Focus on the exciting and you’ll risk missing the useful. Data infrastructure is likely to be boring.
Coverage and impacts
If you take a look at this map of all of the EA LIDAR data you can see that it’s not a complete map of the UK. The data is collected for EA’s operational needs so the choice of areas to scan are based on what they need to know. I’ve seen a few people disappointed that their home city or area isn’t covered.
However there are lots of commercial firms that have offered LIDAR surveys for many years. I recently chatted with someone that had used a commercial survey to fill in coverage in an area of London, as part of a solar installation project. So it’s possible to fill in the gaps, at a price.
I think it would be interesting to explore whether:
- the EA LIDAR release has impacted those businesses, or has it just been a useful sales and marketing tool to help showcase the value of LIDAR surveys?
- are customers of LIDAR surveys interested in just the raw point cloud and/or surface models as released by EA or are there value added services to be offered?
- how do the typical surveys they do compare with the area covered by the EA?
- what are the typical costs for surveying an area to a similar resolution as the EA use?
Collaborative curation as a pattern
The reason why I think this is interesting is that if the EA data is genuinely useful, then perhaps there should be a national dataset that covered the entire UK? It’s unlikely that the EA would want to take on this responsibility as it’s likely to be outside of their remit. So what would it take to fill in the gaps? Is there a collaborative business model that could be explored here?
It would be interesting to consider whether there’s a business model for extending the LIDAR survey coverage that might include:
- shared costs for regular basic surveying and data release, including data collection, analysis and publishing
- value-added services for data analysis
- value-added services for more frequent or detailed surveys
I think the Ordnance Survey may also do some LIDAR surveying or at least benefit from it. So is this an area they could invest in? If the data is useful in a variety of businesses, then sharing costs might allow for a more comprehensive national dataset without negatively impacting existing businesses offering services around LIDAR data.
Taking a step back I wonder whether this is a pattern that we can expect to reoccur in other sectors. Specifically:
- an open data release highlights the utility of wider availability and consumption of a specific operational dataset
- the dataset does not have the ideal coverage, is not as up to date as it could be, requiring further investment
- organisations in a sector come together to explore benefits of a shared ownership model
The other examples of this that I’m familiar with are the model used by legislation.gov.uk and organisations like CrossRef and ORCID. Although in each case the motivations and interactions are slightly different. I’ve previously described a way to characterise these projects, if you’re interested.
I also wonder whether we might see similar activities emerge around national statistics and other survey data. For example what might we learn if the Defra Family Food Survey was more regular and comprehensive?
As I’ve suggested previously, I think what we need is a data infrastructure incubator to help explore these ideas and begin brokering interactions within a sector.