UKH

I have been looking at apps and have found that if it provided GPS data and grid reference it's six digits for the grid reference

I wondered if there was anything better for GPS location, providing grid reference within 1 m that was free.

It needs to be accurate for finding specific points on a walk

In reply to chalk bunny:

OS locate will give 10 figure grid references

In reply to chalk bunny:

As I understand it the main US based GPS systems are designed for the military and can provide it with even better accuracy than you need. However access for everyone else (us) is desensitised to not provide such accuracy, in case it be misused by those inclined to use it against the Interests of the US and allies.

My Garmin watch provides an 8 figure grid ref, I assume that is as good as it gets. Locations with lots more digits may not provide commensurate accuracy?

of course my understanding may be totally wrong 

In reply to kevin stephens:

It seems that I was out of date as the desensitisation was removed by Clinton in 2000. However the hardware of phones etc limits resolution to around 5m. For much greater resolution you would need to invest in specialist surveying equipment which works on a different band

https://en.m.wikipedia.org/wiki/Global_Positioning_System#:~:text=Consumer%....

In reply to chalk bunny:

I use this for getting grid refs & postcodes etc:

https://gridreferencefinder.com/

In reply to ChrisJD:

I have used this to look at location of huts to outddor trips, its a good tool

In reply to probablylost:

I have downloaded this and with have a look- just found the seeting to alter it. Thanks

In reply to chalk bunny:

What3words can display OS 8 figure grid alongside the 3 words.

In reply to probablylost:

Another recommendation for this. Its free and it works.

In reply to kevin stephens:

The US reduced the accuracy during the Balkan war. A side effect was that lots of unrelated things in random EU countries failed. Navigating barges down the large rivers, for instance, infrastructure projects etc. So the EU made their own system, Galileo.  The Russians made their own system, called Glonass. A lot of spots watches use two or more systems at once, to improve accuracy and coverage.

Obscure trivia. The internet uses GPS satellite clocks as a time source, NTP stratum 0, These time signals then get cascaded down, to other devices, including your computer.

In reply to chalk bunny:

"Grid Reference" by Arthur Embleton.

Android & iOS [edit: looks like Android only]. Small. Free. Does exactly what it says on the tin. Can share GR, too.

Will give 10 figure (1m) GRs, but your GNSS receiver won't give much better than 3m accuracy, on a good day.

https://play.google.com/store/apps/details?id=net.blerg&hl=en_GB&gl...

In reply to captain paranoia:

^ this is the correct answer

In reply to ChrisJD:

A great tool - used itfor checking all the grid refs in the Culm Guide,

In reply to dread-i:

Even more obscure trivia - Lighthouses in the UK and Republic of Ireland use GPS time stamps to synchronise their communication system.

In reply to chalk bunny:

The cynic in me suspects that no GPS can give you a grid reference to one meter accuracy as the spheroid projection isn't that accurate - if you compare a point at Penzance to a point at John o' Groats, both lat longs are converted to points on an XY grid, and that requires some stretching in the south and compression in the north.  The consequence is normally decreasing accuracy away from the defining meridian, and you'd be surprised how fast it builds up, and an grid ref xy no longer defines a unique point.

In reply to wbo2:

https://xkcd.com/2170/

In reply to captain paranoia:

As stated by Longsufferingropeholder, that is the correct answer, though only for Android.

For iOS, use GPS OS Grid Reference by Tom's Trails. 

Both are ridiculously simple and provide the grid reference big and bold and unembellished which is the way you want it on those occasions when you REALLY need it.

In reply to dread-i:

If you want to view the British National Grid references in the app, open the menu and tap display settings and then select Country-specific systems and the format you’d like to use. The preview box will then show you how this appears.

https://support.what3words.com/en/articles/4152939-what3words-vs-gps-coordi...

Just double checked and it can do 6, 8, or 10 digit grid ref.

From personal experience, when calling for an ambulance they will ask for w3w. MR will ask for w3w and also grid ref. Because of tree cover, I had one app saying 'you must be outside' whilst another app worked ok. Not certain how that occurred as they both use the same hardware.

In reply to dread-i:

On a tangent, but it really hacks me off that W3W, a commercial company, are monopolising dealing with the emergency services when it comes to location finding.

In reply to Ridge:

100% agree with you on that one.

Although after reluctantly downloading the app on insistence of a friend who works in a 999 callcentre, I educated him on the fact that the app also works with grid references - previously he'd told me they had no way to use grid references, which I found rather shocking.

Having played at home it can easily be 20m off my actual location, so not convinced with its claim of accurately.

In reply to Ridge:

....and the maps they use with it are rubbish. I had firsthand experience of this in the summer when some of my scouts got lost and called 112 🙄.  Helicopter found them and sent a w3w reference to the Police on the ground, who then blundered through woods and bogs in the middle of the New Forest because they were essentially using googlemaps which didn't have all the paths marked on them.  It was pretty amateurish to be honest with you.

In reply to chalk bunny:

GPSstatus app for android will give you 8 figure OS references amongst a whole load of options.  It also tells you what satellites it can see and how good your accuracy is. Very useful for geocaching and checking why other apps using GPS are not working properly!

In reply to Jenny C:

That's probably just that your phone doesn't have a good fix on your location rather than being a fault with What3Words. I'd wager another app giving a grid reference would have been off by similar amounts. The W3W grid is apparently about 3m, which is probably as accurate as you could hope for from a phone GPS fix anyway.

In reply to Ridge:

I'm backing https://www.fourkingmaps.co.uk to take over

In reply to Luke90:

My point though is what's the point of W3W promoting themselves based on 3m accuracy, when in a real life situation phones frequently don't have that good a fix on satellites?

In reply to Jenny C:

Because people (and organisations) will buy into the hype.

In reply to chalk bunny: If you want some fun with maths, consider the stretch that goes into the OS map,,, at Lands End one degree of longitude is 71kms, at John o'Groats it is 58kms, yet these are stretched to be equivalent on an OS map. What inaccuracy does that introduce, or is OS CRS doing something clever to keep it's northings parallel yet reduce stretch.

Bizarrely that would mean with 3m tiles W3W is more precise than OS can ever be , no matter how many digits a grid reference you use as the error/uncertainty would exceed that

In reply to wbo2:

That's the whole point of a local (national grid); you can optimise the projection to minimise error.

It's also why GMA varies across the OSGB coverage; whilst magnetic variation varies with the earth's magnetic field, the projection error (grid divergence) is fixed, but depends on your position within the mapped area.

In reply to wbo2:

https://commons.wikimedia.org/wiki/File:Ordnance_Survey_National_Grid.svg#/...

In reply to Jenny C:

Their GIS may not be able to use OSGB GRs, even if W3W app can present them.

The fact that our emergency services cannot accept references in our national mapping system, that has been around since 1936, is a bloody disgrace, given that almost all commercial mapping systems can (e.g. StreetView, which has been doing it for more than two decades).

In reply to captain paranoia:

You'd at least think the call handlers would have the tools necessary (Google) to translate them

In reply to Jenny C:

Of all the reasons to disapprove of W3W, I don't find that one particularly compelling. As far as I can tell, their system is genuinely entirely capable of giving location references to the precision they claim and it doesn't seem fair to me to hold them responsible for the various reasons the location reported by the phone might be inaccurate, it's not under their control at all and it would impact any other method of getting a location from a phone equally as much. Perhaps if they were claiming a precision that phones were entirely incapable of providing and would therefore almost never be available from their app, I'd consider that a bit cheeky, but phones often can get to that kind of ballpark when they're outdoors with a decent view of the sky.

In reply to Longsufferingropeholder:

You'd think. But no.

In reply to dread-i:

It is true for both postcodes and w3w that they are not open source systems and no comprehensive hard copy mapping is available.  Postcodes are proprietary to Royal Mail Group and are continuously but almost imperceptibly changing (as well as sometimes being 10km across). w3w is also a proprietary system, that relies on the software of the owner and could disappear in an instant. It exists in about 78 languages: do you operate in the language of the caller or the language of the control room operator? 

The base system for position systems in the emergency services in the UK, and now dozens of countries across the world, is good old Lat-Long in decimals of degrees. That is because AML, a system invented by BT to enhance the police's EISEC, automatically harvests mobile phone GNSS data, and other location sources like base station info, during a 999/112 call and sends a datastream to the Public Safety Access Point (PSAP) which is the BT999 operator. The data is then available to be forwarded to the subsequently-connected control room. The data includes latitude, longitude, radius of position (error radius), time of position, plus some techie phone details. 

This data is available everywhere in the UK but not all emergency control rooms have set up their systems to handle it. I deal with Police Scotland who are set up for AML. Outputs coming from their control room to MR will often be in the 12-numeric full national grid reference which can confuse some people but it is on all paper maps as well as many online resources and mobile apps. 

What most people do not realise is that one of the biggest problems for hill-goers and rescuers alike during an emergency is that not many bad things happen in wild and mountainous country. Pretty much every emergency control room in the country is focussed on bad things that happen in towns and cities 99+% of the time. There are more landlines (on EISEC), more base stations, more postcodes, more wifi, more street signs and so on, which provide the data that creates the pattern of their normal day. Along comes a broken ankle on a mountain or a car crash on an barely-used unclassified single track road, and oops, the normal pattern doesn't fit. My view is that you should force the grid reference on them because it is what the folks at the sharp end will eventually need. 

See the EENA website for more about AML:  https://eena.org/ 

For some reason, perhaps another feature of a failing nation state, only very poor information about the status of AML in the UK is available in 2023. However, in 2020, we know that 75% of emergency call from mobiles provided AM data. We can expect that to have increased as older phones get replaced. 

In reply to Luke90:

Accuracy of a fix depends if the device depends on the number of GNSS constellations the device enabled for, how much sky it is exposed to, whether A-GPS or EGNOS are enabled, and a number of other factors. There are two EGNOS ground stations in the UK and one recent report showed a Galileo fix of 0.860m across based on the Glasgow one. 

The discontinuity of w3w is a real pain. If a pperson needing help was originally at NG95321357 and now they are at NG95311357 then it's obvious they are essentially in the same place but a few steps to the West. With w3w, the person simply turns around on the same spot and their phone reports something completely different.  

In reply to Jim Fraser:

Slight deviation, but on postcodes, in case you ever need it this is super useful: https://postcodes.io/

In reply to Jim Fraser:

It could do, though their plan is to continue to throw money at marketing it (hence the hype) until they achieve a certain amount of what I think is called 'market penetration', and then once enough people and organisations are dependent on their proprietary system - ker ching!!

Here's an excellent little video that always needs posting when W3W is mentioned:
youtube.com/watch?v=SqK0ciE0rto&

(To reply to your following post as well.) 
An alternative system has existed for years that has most of the pros and few of the cons of W3W:  https://en.wikipedia.org/wiki/Open_Location_Code
 

It was developed by Google, but it's open source so nobody (including Google) can monopolise it in the future.  As such, Google's apps and maps use it natively, so far more people already have this functionality on their smartphone than have W3W even after years of their aggressive hype.

To identify your current location to a W3W level of precision using Google Maps, open up the app, press the little blue dot that marks your current location and the 'your location' window opens up, then scroll down.  (The app will warn you if the GPS location it's giving you is sketchy.)

If your phone doesn't have data it'll display the full global OLC code (11 characters and the '+' sign) if it does it'll miss out the first four characters and add a placename at the end to identify which 40km (ish) square you're in.

11 characters seems like a lot compared to 3 words, but the character set is specifically chosen so that it's difficult to confuse any one character with any other and it only seems like more to deal with until you start having to spell the words out because they're ambiguous.  Or you start running into problems because someone isn't working in their native language, has a strong regional accent or something, or is just struggling to spell or pronounce a word and misunderstandings start creeping in.

For example, if I were standing in the entrance to the Awesome Walls car park in Sheffield, I'd be at:  9HX3+WMM Sheffield

To find the 'plus code' for a location other than where you are using the Google Map app drop a pin and scroll down through the 'dropped pin' bit.  Or click an existing pin and on the 'overview' tab click 'see all'.

In reply to captain paranoia:  Thsnks - I'm use to numeric only reference systems, so it's a clever trick to basically chop the UK into lots of little squares, each with it's own CRS... does the ESRI GIS (industry standard) handle that?

In reply to captain paranoia:

A good illustration of the fact the precision and accuracy are not the same thing.

In reply to wbo2:

No; the coordinate reference and projection is the same across the entire mapped GB area. That's why grid divergence varies across the UK.

UTM divides the globe into zones in the manner you describe, which is why you need to specify UTM zone when quoting a lat/long, and zones overlap, and the same lat/long in different zones point to different locations...

In reply to captain paranoia:

This OS document probably tells you more than you will ever want to know...

https://www.ordnancesurvey.co.uk/documents/resources/guide-coordinate-syste...

In reply to wbo2:

The OSTN15 transformation used by the OS to convert from GPS coordinates to OS Grid coordinates is a polynomial grid transformation and not a spheroidal projection. It has an average error of 0.1m (obviously not counting the error in the GPS coordinates in the first place)

If I were implementing this in an app or smartwatch, an easier method is a Helmert transformation which is accurate to about 3m.

In reply to planetmarshall:

UKC at its finest. All the OP wanted was an app to tell them where they were, and here we are discussing the pros and cons of various projection transformations...

In reply to deepsoup:

The other big issue with w3w is that if you accidentally use a plural when your location is singular or the other way round then it could put you somewhere miles away from where you actually are let alone if a word is misheard.....  At least with OS Grid if a character is misheard it still gives you a location fairly close to where you are

In reply to captain paranoia:

Mention of UTM leads us to an interesting historical feature of the OS grid reference. During WW2, the allies, notably the USA, noticed that British military effectiveness was substantially enhanced by the ridiculously simple means of knowing where you are and knowing where you are going. This was facilitated by the 'British System' that had morphed into the 'Modified British System' or 'Cassini Grid'.

Like the current OS Grid Reference System, this divided areas into 500km and then 100km squares designated by two letters such a qY. Within the 100km square a numeric reference was used, just as today. This enabled soldiers to designate locations locally with 4 or 6-figure references, and so on, in a very simple manner. 

UTM's origins and development are swathed in the mists of time but a working system was in place during WW2 and may have developed during the 1930s with German and American influences. After WW2 the Americans set about creating a system as simple for the end-user as the British Cassini Grid. That became the original version of MGRS (Military Grid Reference System) and is probably connected with a couple of other military coordinate regimes I semi-remember from ancient times. Between the mid-1980s and 2000, it further developed into the NATO-wide system that exists today. In 2001, that also became USNG (US National Grid). 

Within MGRS/UTM Zone 30V, the top of Ben Nevis is in 100m square UH776965, while in the OS Grid it is in 100m square NN166712. Though the letters and numbers are different, the format, and the methods of derivation and use, are identical, though based on different ellipsoids and meridians. The now widespread adoption of MGRS is a powerful indicator for the simplicity and utility of the referencing format developed by Ordnance Survey about a century ago and still in use today.

I commend this motion to the House.

In reply to deepsoup:

Thanks - hadn't seen that before.  I can see that the 'last mile' argument for places with poor/no street addressing is a valid use case, but that's about it.  The commercial and closed nature of W3W is a problem for me, trying to monetise an emergency location service, when adequate free alternatives exist.   

This is an interesting read:  https://cybergibbons.com/security-2/why-what3words-is-not-suitable-for-safe...

I was surprised to discover that 'mayday' is on the W3W wordlist, and indeed that mayday.mayday.mayday is a valid location.  

EDIT to add this: https://keswickmrt.org.uk/rescue/crow-park-keswick-83/, from https://w3w.me.ss/ which is, um, interesting.

And edit to add: I have just invented a W3W game - put in three words you might use in an emergency and see where you are!  cold.broken.legs

In reply to captain paranoia: I think it's useful as you often see degrees of location accuracy i.e. my gps is accurate to 5cms, w3w is no good as it's only accurate to 3m , without any thought applied to how accurate maps and locations are really, when you get to the nitty gritty of how coordinates are generated.  Most of the time I work with UTM, and that can get pretty inaccurate (30m+ - all points point to the same lat/lon) when you get far enough from the central meridian, and at the point where you'd be considering flipping to the next zone

As stated there's a difference between precision and accuracy, and I know a bit more about National Grid, so I've benefited.

In reply to wbo2:

Pushing it a bit, even for a differential setup with local reference station and centimetric wave receiver...

IIRC, the GSVF-2 simulator we developed for ESA had an accuracy of 2mm in the signal it generated...

In reply to SNC:

I think that use case is one where the Google 'Plus Code' (Open Location Code) is clearly superior.  Their philosophy of deliberately avoiding words (in any language) and using a reduced character set designed to avoid confusion as much as possible seems much better suited to that.  As does the existing near-ubiquity of the Google Maps app. 

But the biggest advantage would seem to be that it's open source.  So if a postal service in a developing country (say) wants to use it they can and it won't cost them anything, ever.  And if they want to develop their own software or their own app to deal with those 'addresses' (eg: to help plan delivery rounds and manage logistics), they're completely free to do so.

Crikey.  I wonder if that was one of those examples of a plural word being plausibly close but inconveniently far from its singular.

In reply to captain paranoia:

All this centmetre and millimetre surveying stuff makes me laugh. The Eurasian Plate is moving at 7 to 14mm per year!  

In reply to SNC:

A 999/112 call (or SMS) from a recent (Android 2016, Apple 2018) smartphone should provide a valid AML datastream. Based on 2020 data provided to EENA, and making an estimate for the uptake of AML/ELS capable phones, one might reasonably expect between 80 and 90% of emergency calls from mobiles to result in a good AML fix and if the caller is outdoors then that is usually from the phone's GNSS receiver. There are still control rooms out there that have not adapted to AML and cannot use it. For instance, in 2020, EENA recorded that only 60% of police forces were using AML.  

In reply to Jim Fraser:

I think you misunderstand; GSVF-2 was a constellation simulator; it generated the signal-in-space for a receiver. It was the 'gold standard' for testing receivers. The error spec allowed ESA to evaluate the performance of Galileo receivers, and know the additional error added by the receiver. In a purely synthetic, simulated space.

In reply to SNC:

Excellent game thank you. I wouldn’t want to be stuck at exploit.public.data without having paid my W3W subscription - looks quite a long way from anywhere. 

In reply to deepsoup:

 .... Google 'Plus Code' (Open Location Code) is clearly superior ....

Agree totally.

But from talking to people, w3w is, I think, more intuitive for the average person (not saying it's better).  

For some real computational nerdery:  https://arxiv.org/pdf/2308.16025.pdf 'A Critical Analysis of the What3Words Geocoding Algorithm' by Rudy Arthur, Uni of Exeter.

In reply to deepsoup:

W3W use Google locations to give their magic code so it is much better to use Google for location data and cut out the commercial middle man.

In reply to SNC:

They'll be warming up the lawyers...