In October 2013 the OfCom infrastructure report detailed the state of coverage in residences and populated areas but also emphasised that ‘mobile coverage away from the home is also important, especially on roads and rail. (p.6)’ Mobile coverage on these transport networks is increasingly important, with mobile connectivity adopting a central role for in-car navigation and entertainment systems (the rise of the ‘connected car’) and for enabling remote working while travelling on railways.
As connected cars become more sophisticated, and more data-demanding, 3G/4G coverage on roads will only become increasingly important. This report looks at UK Motorways, A-roads and Railways - click on the tabs below to see how the networks perform for these different categories. Hover over the roads on the map to see the specific stats for that section of road or railway.
For each road we provide a connectivity summary, the aggregate by-network stats are visible below. We get our road shapefiles from Ordnance Survey (so they’re the OS to blame for any mistakes, not OpenSignal!) and we are reliant on their labelling. In order to make sure data is from a user driving down that road, we make sure the data points come from the road area and from a device that has travelled along that road for a significant distance - this helps us to eradicate unreliable points or those that might come from people who are not in vehicles.
The OfCom infrastructure report confirmed a commitment to studying coverage on roads, but the statistics used by OfCom were based on modelled projections - and focussed on whether roads were served by none or all of the networks. They indicated they intended to do more intensive work on this subject in 2014 - so this report gives the best current overview of mobile connectivity on UK roads.
Mobiles are already being used in cars for navigation, to stream music - and increasingly to run in-car operating systems that themselves depend on mobile connectivity. Connected cars will play a huge part in the future of the Automobile industry, and mobile connectivity on Motorways and A-roads will be a vital part of the infrastructure required to make them an effective reality.
Time on 3G/4G is a metric we have developed that looks at coverage as it is experienced by users, rather than by pure geographical reach. For this report, it shows what percentage of time users have access to 3G/4G technology when on one of the selected roads. What this means in practice is that an area of road that typically has more people driving on it will be weighted more heavily in the overall score, because we will have more readings from there. This is, however, reflective of demand, as it is more important to have 3G/4G coverage where it will be of use to more drivers.
For Motorways in the UK we see that the average coverage is 76%, meaning that drivers have access to 3G/4G 76% of the time. The best performing network for this metric is 3, which serve their subscribers with 3g/4g coverage 92% of the time. At the other end of the spectrum, Vodafone and O2 are very close - with their users experiencing 65% and 67% time on 3G/4G respectively.
The OpenSignal app checks to see connection type approximately every 10 minutes, which is how we are able to show accurate ‘time on network type’ stats. In the event of no coverage (or emergency calls only) the app records the GPS position of the phone and logs that no network has been detected, which is how we are able to show ‘Time on No Signal’ stats. This can also be understood as a proxy for understanding total coverage (100% -time on no signal), EE, the best network for this metric, therefore have 95.4% 2G/3G/4G coverage on UK Motorways.
As can be seen from the graph - the networks are all extremely close, with a spread of only 0.3% between EE and Vodafone/3. With the average time on no signal below 5% for all networks we can see that, at the very least, 2G is available most of the time on UK motorways.
The time on 3G/4G metric doesn't tell the whole story, however. The UK’s 4G rollout has meant that mobile networks are now capable of delivering extremely fast performance to mobile devices - something which will only be more important as connected car systems become more sophisticated and therefore more dependent on advanced mobile technologies. Also, owing to the propagation qualities of the 4G LTE spectrum, 4G is able to travel much further than 3G signal - meaning that fewer towers would be needed to cover the same area of Motorway.
This metric also helps to clarify the 3G/4G metric above - 3 have the best overall network for 3G/4G coverage on roads, but their time on LTE is the lowest at 21.3% - showing that their coverage is overwhelmingly 3G. Systems (such as in-car video streaming for passengers) that rely on 4G may well become more common, meaning that 4G coverage on roads becomes a higher priority. EE perform very well for this metric, with over 50% 4G coverage on UK motorways.
Time on 3G/4G is lower for all networks on A-roads than it is for motorways, with the average falling from 76% to 67%. This supports the findings of the OfCom infrastructure report, which says that ‘The data shows that coverage on motorways is good, but that there are significant gaps in coverage on A and B roads, particularly for data services.’ That being said, we show, for A-roads at least, that there is still significant 3G/4G coverage.
On UK A-Roads, overall coverage is relatively similar across all networks, with a spread of only 0.8% between the networks. Again, the data shows that coverage on A-roads is worse than on Motorways - with the average network user being without signal 6.6% of the time when driving on an A-road, compared to 4.8% on Motorways.
For A-roads we actually see the average 'time on 4G' as slightly better than on motorways, with a small improvement from 36.8% to 37%. For the other two metrics we see Motorways as performing better, so this is slightly surprising - but perhaps explainable once one considers the nature of the 4G rollout in the UK. Unlike Motorways, A-roads also go through cities - and these urban A-roads are likely to be some of the most highly traversed, leading to more readings from within cities than rural areas. With 4G rollouts usually focussing on cities first, this might explain why both Vodafone and O2 users have better 'time on 4G' stats on A-roads than motorways, and why EE - who are at a more advanced rollout stage - perform worse (as might be expected given opposite findings of 'time on no signal' and 'time on 3G/4G' metrics).
Connectivity on trains is especially important for business travellers and commuters, allowing them to work without being in the office. Reliable cellular internet on trains will make it much more seamless for workers to divide their time between UK offices and will lower the opportunity cost of travelling between them. Connectivity on trains is also important because I want to be able to watch youtube on particularly boring stretches of the journey.
For the map we detail how many signal readings we have had from each stretch of track in the last 3 months. If the number of tests is below 100 we make that stretch of track clear, rather than showing a colour, as there is probably not enough data for us to be certain that it is fully representative (this is because some of the stretches of track are quite small, as the shapefiles we use are sections of track - rather than the track as a whole). Hovering over that section will still show the data we do have, but it may not be completely accurate.
UK 3G/4G coverage on railways is better than it is on A-roads but worse than it is on Motorways. The average rail user has a connection 72% of the time, while a user on a Motorway has a 3G/4G connection 76% of the time and a user on A-roads having a connection 67% of the time. Coverage on railways was something that was flagged up as especially tricky to measure in the OfCom report, owing to the impact that the trains themselves have on mobile signal: 'Even where track side signal is strong, some types of train carriage can significantly degrade the signal and prevent reliable service' (p.49). The only reliable way to measure the true state of coverage as it is experienced by users actually inside train carriages is to measure at point-of-use, as we do.
The average time on no signal of someone on the UK railways is 6%, with EE performing best for this metric at 5.7% and Vodafone performing worst at 6.4%. However, it would seem reasonable to assume that on trains it is the data connection that is more important than voice - and so this metric does not give much more of an insight into how mobile networks work for the poeple that use them than time on 3G/4G.
As more commuters work on their journeys (a trend noted by recruitment firm randstad) - 4G coverage becomes more important, helping to minimize the difference between data speeds on the move and in the office. We again see EE performing best for this metric on trains, with the average EE user connected to 4G over 50% of the time.