This quarter’s quarterly project is “Food Hygiene Ratings”.
See the 2026 Q1 Project wiki page for more details.
Feel free to add useful stuff there, and to ask questions and discuss things in this thread.
(Edited to add: In the absence of any other community discussion or engagement by the time the Quarter had started, a topic was picked by OSM UK Board at it’s monthly meeting on 6th January.)
The current UK quarterly project wiki page says:
But the forum search doesn’t seem to show that thread. Where was the vote for this?
Perhaps add a worked example here explaining a typical activity from https://gregrs.dev.openstreetmap.org/fhodot/ ? Not everyone reading this will be familiar with it.
Thanks Robert. I’ve certainly done this a lot in the past and we haven’t had this as a quarterly project in a while, it might be the motivation improve my area.
There’s a few suggestions on the QP wiki page linked in the first post of this thread. If anyone has any other ideas or wants to add any more detail there, then please do.
Indeed - more detail is what is required - a worked example would be great.
I’ve started working on updating my local area and I’ve got to say, using the the FHRS/OSM comparison tool has been really easy to sort out the suggested matches with the JOSM integration.
I’ve had a couple of discoveries that have raised some questions for me though:
For better location accuracy, there are three local authorities which use UPRNs in the LocalAuthorityBusinessID FHRS field (fhrs:local_authority_id): Castle Point, Fareham and Vale of White Horse. It doesn’t apply to all businesses in those areas, as identifiers from older schemes are also used. In Fareham’s case, a suffix like A or B is sometimes added to the UPRN.
For mobile caterers where there’s likely no on-the-ground evidence of the business being based at the address, and the public facing parts (i.e. the mobile van) operates elsewhere, I don’t think it’s correct to add the FHRS ID to the address.
For your first example, I think it depends. If a public-facing business is being operated from a residential address, then arguably it could be mapped in OSM, and then it would be appropriate to map the FHRS ID there too. (Ideally I think you’d have a separate node for the business, inside the way for the house itself. However, if the business isn’t publicly advertised, then there’s a privacy issue we should probably consider too. It’s possible their name and address has been included in the FHRS listings by mistake. (There’s an opt-out system on the public register for private residential addresses.) So I’d probably either survey the address on the ground, or look for a business website that advertises the location before adding the business and FHRS ID from the FHRS data.
It seems in Derby there is an aim to get 100% of FHRS IDs mapped, even if there is no signage on the ground showing that the business is at that address. I agree that it doesn’t seem right to do so.
Thanks. I thought that would be the case. Think I’ll put in an issue on the FHODOT tool to explore potential ways of having this data excluded from the tool as an exemption from mapping.
Cool,
We have a similar system for Denmark.
Almost all restaurants, cafes, bars, bakeries etc are linked to a Food Hygiene report.
Are people adding businesses to OSM based on FHRS data?
I thought the idea was that we add FHRS IDs, postcodes etc. to businesses that have already been mapped, based on surveys. In that case there is no need to exclude certain types of businesses from the matching.
Adding businesses to OSM that are in the FHRS data but not in OSM wouldn’t be a good idea in my area, there are too many outdated entries for businesses that no longer exist.
It would be nice if FHODOT had a button I can click to report this, so it’s not included in the stats, and it doesn’t keep getting shown.
I have certainty done so on occasion. From a legal/licensing standpoint, I think it’s absolutely fine to do so. But I would want to be reasonably confident that (a) I can identify the correct location for the business and (b) it is still operating in that location.
The location can be fine if you have a regular street address in FHRS and the address is already mapped in OSM (perhaps as part of a previous business in the same property) or its location can be inferred by interpolation. This works well on typical high streets. It’s less good for more rural locations.
Knowing whether the business is still operating can be more difficult. There’s a " Date of inspection" date included in the FHRS data. If that’s reasonably recent (e.g. within the last year or two), then there’s a good chance the business will still be there. How long in the past you think is acceptable might depend a bit on the type of business. For independent businesses you might be able to find a website for the business that adds weight to it still being there.
If there’s a business already mapped in OSM, and there’s a FHRS listing for that at the same address/location with an inspection date that post-dates when that OSM object was last edited, that may well tell you that the currently mapped business must have been replaced. Even if you’re not completely sure if the new business is still there, it would still be an improvement to the map to replace the old business with the new one. (Currently it’s definitely wrong, if you make the change, it might not be wrong.) If the inspection date was a while ago, you could add survey:date=* for the inspection date, which lets other editors (and QA tools) know that the change might not be current.
I’ve just been looking at Survey Me in my local area, turns out a bunch I had previously mapped were flagged as “Update needed for FHRS ID mismatch” AKA the entry in FHRS no longer exists, so the OSM link is broken. I checked them out:
check_date and removed the old fhrs:id.Sheffield’s matched percentage is up from ~48% to ~60%. A lot of the unmatched FHRS data seems to be because of out-of-date POIs, especially in the east and north of the city and in various shopping centres.