You probably don’t realize it, but you’re a direct beneficiary of OpenStreetMap. You may not be one of those who has contributed the 100 million edits to the community-driven mapping project, but if you’re a customer of Apple, Facebook, Uber, Amazon, or countless other organizations, you depend upon OpenStreetMap every day.
For example, I mostly interact with OpenStreetMap when using Strava, which builds on Mapbox’s OpenStreetMap-driven data. Back in 2014 I asked if OpenStreetMap could become the next Linux. Fast forward to 2021 and the number of contributors has more than doubled, even as the number of organizations depending on its open mapping data has skyrocketed.
So maybe that’s a “Yes”? But challenges remain.
OpenStreetMap by the numbers
First, let’s look at some numbers. In mid-2020 Accenture published a report that found the value of OpenStreetMap’s data topped $1.67 billion. (This is roughly how much the Linux kernel was valued at back in 2008.) That’s a lot of mapping data available for free. The reason? So many people and organizations give their time and talents to it:
- As of August 2020, OpenStreetMap saw 4.5 million map changes (edits/changesets) per day;
- Since its inception in 2004, people have contributed more than 100 million changesets (that’s an individual choosing to use their talents 100 million times to improve mapping data for all);
- This equates to nearly 1 billion features globally in the past 16 years;
- As of late December 2020, there were 721,270,948 “ways” (a linear feature on the ground, such as a road, wall, or river) within OpenStreetMap’s data;
- Roughly 50,000 contributors participate in OpenStreetMap each month; and
- Since the start of the project, over 1.5 million people have contributed.
One of those contributors, interestingly (at least, to me!) is my brother, Clark. He’s now a law professor, but over a decade ago he was active with the OpenStreetMap Foundation, providing legal assistance. Yes, even lawyers can be productive contributors!
OpenStreetMap is a colossal success story, one that keeps pressure on commercial mapping data providers like Google. (It’s worth noting, however, that OpenStreetMap isn’t a Google Maps competitor. OpenStreetMap offers mapping data, not a consumer-facing mapping product.) Not only does it force other providers to keep costs low, but it also puts pressure on others to cover areas of the world that they might otherwise skip. And yet OpenStreetMap still has plenty of hurdles to overcome, balanced by robust strengths.
OpenStreetMap strengths and weaknesses
- Huge range of what is mapped (not just streets) due to its proven data model with built-in flexibility and freedom in what can be mapped.
- It’s homegrown, made by pedestrians and cyclists and so favors details (e.g. the design and contents of a park, rather than how to drive past it fastest).
- Immediate feedback from making an edit to seeing it on the map.
- Tools for contributors are largely available in a fairly broad spectrum of languages, presenting relatively low language barriers to contributors.
- A place to find detailed domain knowledge and lively debate over how to map anything, coupled with a committed community.
As for weaknesses, well, let’s just say the OpenStreetMap community is far harder on itself than it should be. The list of weaknesses goes on… and on… and on. But a few stand out:
- OpenStreetMap attracts a lot of map contributors, but needs more help in its technical foundations. For example, the project lacks a database administrator to refactor the main database and replication API to make it easier for consumers to consume diffs. Plus OpenStreetMap apparently has a weak data maintenance culture, biased toward dwindling greenfield mapping areas.
- Though the community has grown considerably, contributors complain of a lack of inclusivity at times, and not always the most welcoming of “vibes.”
- As OpenStreetMap grows, it risks becoming too complicated (“More and more parts of our map can no longer be edited or maintained by normal users.”)
These are real weaknesses, presenting real challenges, and yet the opportunities outweigh the impediments. Location data is simply too important, too foundational, to be owned by any particular company. It needs to be community-driven.
Like Linux, OpenStreetMap increasingly draws on individuals who are paid by their employers to make OpenStreetMap better. Some of the technical problems with the project could be overcome if DBAs, for example, were paid to improve the underlying infrastructure. This sort of involvement will change the nature of OpenStreetMap’s community, but isn’t necessarily a bad thing, just as it wasn’t for Linux. Some of the more prominent concerns with the OpenStreetMap community (diversity, inclusivity) would be aided, not hurt, by involving more corporations that demand such attributes in the open source communities to which they contribute.
The tl;dr? Yes, OpenStreetMaps can be like Linux. But today it may be like Linux was in 2008. To get its next 100 million edits, it may be time for corporate interests to move beyond using OpenStreetMap data and contribute more to making it better.