Data sources and freshness
Which open sources OVatlas uses, how fresh each part of the map is, and what the site does and does not keep about you.
The sources, listed
OVatlas invents nothing of its own; everything on the map comes from the open data of Dutch public transport. Vehicle positions and disruption reports come from the NDOV loket, the national counter for open transit data. Timetables, stops and routes come from OVapi. Shared rental vehicles — hire bikes, scooters and the like, in the cities where they are available — come from Data Deel Mobiliteit. The same list appears in the privacy policy, because it is also the list of parties with any involvement in the site.
One source sits outside OVatlas itself: the map background. Your browser fetches those map tiles directly from Carto, which means Carto sees the approximate area you are looking at. That is in the privacy policy too, under the services we use, alongside the DB-IP country database our own server consults and to which your IP address is never sent.
How fresh each part is
The parts do not refresh at the same rate, and that is exactly right. Vehicle positions are the fastest: they arrive the moment a vehicle reports and are on your screen within seconds. When a vehicle stops reporting, its position goes stale and the marker is removed from the map — better nothing than something that is no longer true.
The timetable is a dataset that is renewed periodically rather than streaming continuously. Lines, stops and planned times do not change second by second; they change all at once at a timetable change. Departure times at a stop are a blend of the two: the planned time from the timetable, corrected by the deviation the operator is reporting at that moment.
Disruptions and engineering works arrive as messages written by the operators themselves, at the moment they publish them. Such a message can therefore run ahead of what you see happening on the map, or lag behind it; it is human-written text rather than a measurement.
What we do not adjust
The data is presented essentially as it arrives. We structure it so it fits on a map — a position onto a trip, a trip onto a line, a line onto a stop — but we do not correct times and we do not invent positions to fill gaps. That also means mistakes in the source end up on the map: a trip signed on incorrectly, a vehicle carrying the wrong line number, a delay that is simply wrong.
So we cannot guarantee complete accuracy, and the privacy policy says as much. What we can guarantee is that what you see is what was published. When something goes wrong consistently, the cause is usually in the feed itself — and it is the same information that reaches the operators' own apps.
What is kept about you
The question runs the other way too: what does OVatlas keep about you? No account, no name, no email address — we could not receive them if we wanted to, because there is nothing to type them into. Your preferences, such as language, map theme and favourite stops, live in your browser's local storage and never reach our servers. Your location is used only if you allow it, and is handled entirely by your browser.
What does sit on our servers is technical logs: IP address, timestamps, and details of the request such as method, URL and user agent. Those logs are kept for thirty days and then deleted automatically. For visitor statistics we use Umami, which we host ourselves; it sets no cookies and derives an anonymous, daily-rotating identifier from your IP address instead of storing it. The full account, advertising and your rights under the GDPR included, is in the privacy policy linked at the foot of every page.