Mobile map: Some questions answered

Mobile phone mast

On Monday we launched our mobile mapping experiment which aims to use crowdsourcing (a word John Humphrys did not like one bit when we discussed it on Today!) to work out what sort of coverage we are getting from our phones.

There has been plenty of enthusiasm for the project - but also some questions about its methods and aims. So I thought I'd try to answer some of them.

How many people have downloaded the UK 3G survey app so far?

The latest estimate 36,000, and they are spread right across the UK. - from Cornwall to the Scottish isles.

What kind of data are you now getting?

Each of those users is now sending data back to the project showing where they are, what network they are using and the kind of mobile signal they are getting - HSDPA, UMTS, EDGE, GPRS or none.  It's not recorded all the time - just in short windows in order not to use too much battery power.

What kind of map will that produce?

We hope to able to plot information by operator so that consumers will get a good idea of coverage in their area. The more users we get, the more granular the  data we will be able to plot on the map. Every extra download means another street covered.

The app is draining my phone battery - what can I do?

I too had this experience on the phone I'm using for the survey. Now though, Epitiro, which makes the app, has issued an update which makes it less heavy on the battery. What you can do is switch off GPS, which should sort the problem completely, though it will mean we will get less detailed location data. But Epitiro is promising another update next week which will decrease the level of testing, and save more power even if you leave GPS switched on.

How to join the project

  • You will need an Android handset
  • Download the app below or from Android Marketplace
  • Once downloaded, data will be collected without you having to do anything more
  • If you wish to see what coverage is like in a particular place, simply click on the app
  • The app is free to download
  • It uses very little bandwidth
  • The data is anonymised and neither Epitiro nor the BBC will collate or store any personal data

Why is the app only available for Android phones?

As we explained earlier, other mobile platforms make it much harder to capture the data automatically, without some active input from the user. What's more developing different apps for every operating system would be quite a challenge, so we've gone with Android as the easiest system to use. Given how many downloads we've seen, that seems to be working.

OpenSignalMaps already do this - why aren't you using its app?

Collating all the data and then plotting it on a map is already going to be a challenge for our web team - having more than one dataset would make it impossible in the time available. But we've been talking to OpenSignalMaps, whose Android application you can download here.

Their map plotting mobile phones masts across the UK already looks very impressive, so at the end of the survey period we're going to try to look at their data and maps alongside our research .

Why is the BBC doing this?

While the number of downloads shows how much interest there is in the mapping project, some people are suggesting that this is just not a job the BBC should be doing. Well blame me and other members of the technology team - we've been frustrated by the lack of information about the state of our phone networks and somehow persuaded colleagues that we should do something. Just as the BBC sometimes commissions opinion polls, it seems to me perfectly sensible for us to undertake this kind of survey.

Rory Cellan-Jones Article written by Rory Cellan-Jones Rory Cellan-Jones Technology correspondent

Instant translation – no longer sci-fi

Automated translation is no longer the stuff of sci-fi fiction, since Skype launched a beta version of its Translator service.

Read full article

More on This Story

More from Rory


This entry is now closed for comments

Jump to comments pagination
  • rate this

    Comment number 23.

    Questions asked here are answered in the FAQ of the app (press your menu button!)

    Accesses call data because it also records dropped calls.

    Not skewed by flight mode, although it does impact the graph the app makes.

    Knows your wifi connection to account for reporting 2G even if there is 3G available (your phone generally drops to 2G to save battery)

    No idea about the Vodafone 3G boxes tho?

  • rate this

    Comment number 22.

    Will the app be affected by use of UMA? I fully support this idea as hopefully it will demonstrate how far short the operators coverage actually falls when compared to their claims.

  • rate this

    Comment number 21.

    I'm not convinced that the app is recording much detail of calls and applications running, the statement from Epitiro is only saying that the permission granularity allows this, not that the app is doing it.

    I think that there are more important questions to ask, one being why is the data only being collected up to the end of July? Even with 36,000 people using it, that creates pretty thin data.

  • rate this

    Comment number 20.

    #17: Why would it examine current and recent running tasks unrelated to signal type? Why would it even look? Your interpretation of the statement on the site obviously differs from mine.

    Why would an "anonymous" bandwidth checker need to know the phone and serial number, and the number that a phone is in contact with?

    No specific Security Policy for Epitiro, points to Google!

  • rate this

    Comment number 19.

    Googles data security allows this application can get at location and telephony info without the user. If this app can do it so can other rather less scrupulous applications. The app shouldn't kill the battery either if its done sensibly (say every minute)
    All in all calls into question Androids suitability (iOS also has issues... maybe Symbian shouldn't have been scrapped so quickly)

  • rate this

    Comment number 18.

    How can the app distinguish signals from femtocells (e.g. Vodafone's SureSignal boxes)? Does it even try?

    Femtocells are typically used by people BECAUSE they can't get a good signal at home, so if data from these is included it will badly skew the data sent by people in their homes, showing a good signal when there is actually a particularly poor signal or none at all for others in the same area

  • rate this

    Comment number 17.

    @12: Android only has a limited set of permissions so if you want to retrieve something (such as what signal type the phone has), the permission you need gives you a lot of other things which you may well not need. The statement on the site just indicates this and doesn't mean the app is actually retrieving this information.

  • rate this

    Comment number 16.

    Next generation services (LTE) which are frequently called 4G, will use spectrum in the 800MHz range; this will enable further penetration into buildings (won't be available until 2013 or 2014). Some network operators are calling for a standardised measurement. The project by the BBC may help to name and shame those who are reporting coverage incorrectly, IF they can be easily compared.

  • rate this

    Comment number 15.

    Here's the way I understood the test: BBC will be asking users to use mobile phone app which will record availability of both 2G and 3G services. Then they will then create a clickable map to show the data. The project is in collaboration with Epitiro who have developed the application for the android smart phone. It will measure signal strength only, but no testing of the connection.

  • rate this

    Comment number 14.

    At 2.1GHz for 3G, more signal is blocked by walls than the 900MHz used by some 2G networks. This means that users sometimes find that they can still access mobile networks when inside, but are often using the slower 2G GSM network. A recent ruling has meant that networks which use these lower 900MHz frequencies (O2 and Vodafone) can now use it for 3G services.

  • rate this

    Comment number 13.

    Coverage is the number one issue for consumers. Coverage app will provide the information consumers need to see if 3G services are available & from which mobile operator. One problems with mobile phone signals is that signal strength will drop when impeded by something such as buildings. This is largely due to the frequencies which are used.

  • rate this

    Comment number 12.

    From here:

    This application can determine the phone and serial number of this device, if a call is active, the number called etc. WHY?

    Application retrieves information on current and recent tasks run, may allow malicious apps to discover PRIVATE information about other apps. WHY?

    Answers please Rory.

  • rate this

    Comment number 11.

    You also haven't addressed the issue of survey biais towards those who have enough coverage to warrant investment in an Android phone. Those who don't have coverage are unlikely to have one and so you are not capturing a true picture of under-coverage for people who only have low-end phones.

  • rate this

    Comment number 10.

    Some questions not answered:

    Does the app record time in "flight" mode as no signal? (it seems to).

    Does the app record whether the current mode (GPRS, HSPDA, etc) is because that is the best available, or because the user has selected it (usually GPRS to save battery or because phone has WiFi connection)? (Again, it seems to.)

    Failure to take these into account could skew results.

  • rate this

    Comment number 9.

    i condone that the bbc do this study.

    Nice one Rory.

  • rate this

    Comment number 8.


    The days of testing software to death is long gone. With online updates freely available everyone puts something out earlier than they would if the software and updates where disc based. Along with agile development it's the suck it and see method of creating software.

    It's here to stay I'm afraid.

  • rate this

    Comment number 7.

    @ #2

    More datasets measuring the same thing in a differnt way = more informed picture. If both datasets match then groovy. If they dont then it highlights the fact that opensignalmaps are a folly.

    trusting one data source only is a mistake. This will make everyone more informed and that little bit wiser.
    plus, it give you something to compare to the undoubtedly biased operators coverage maps.

  • rate this

    Comment number 6.

    And the release and licensing of the collected data set?

  • rate this

    Comment number 5.

    How many of those that downloaded the app uninstalled it due to battery drainage issues? This app should have been tested more thoroughly before being released into the wild.

  • rate this

    Comment number 4.

    I wish to apoligise in advance to people who live near me and are on Vodafone. The results might end up looking excellent for my road - but that might be because 3G is generally so poor, I've invested in a femtocell device.


Page 1 of 2



BBC © 2014 The BBC is not responsible for the content of external sites. Read more.

This page is best viewed in an up-to-date web browser with style sheets (CSS) enabled. While you will be able to view the content of this page in your current browser, you will not be able to get the full visual experience. Please consider upgrading your browser software or enabling style sheets (CSS) if you are able to do so.