Support road.cc

Like this site? Help us to make it better.

Smartphones and long rides: The Strategy

There's a lot of discussion around smartphones vs. Garmins for long rides. In one corner, Garminados waffle about 15hr battery life, waterproofness and size, while in the geek corner Mobilistas tout cheap smartphones with a few extra batteries as the way to go, not doubling up on tech as the Garmin crowd carry mobiles anyway.

I present proof positive that a smartphone - even one 4 years old - can do the job. This screenshot is 5 1/2hrs into a 'ride', recording with Strava and using Viewranger for navigation, at 25% battery left. Critically the Viewranger Trip screen has been on all the time, with the nav arrow and other information displayed - meaning it can do a ride of over 7 hours with the display on permanently. Brightness was turned down, but it was still quite visible. Also the phone was connected to 3UK the whole time, mobile data and WiFi turned off.

Noting that if you set the screen to auto-off after say 15 seconds, you can wake it to check direction at an intersection, then it will go back to sleep automatically. This will dramatically improve battery life to about 12hrs in this case, as you see below screen power accounts for nearly half battery usage.

So you have Strava logging, always-on navigation, plus can receive calls/SMS. With the option to turn on mobile data to check email if really needed.

The phone in this case is a Sony Ericsson Xperia Arc, but for £40-ish you could have a Motorola Defy from eBay, rated IP67 waterproof and pretty shockproof to boot. Stem-mount it on £27-worth of Quadlock and you have a quarter-turn easy on-off solution that is weatherproof and can have its always-on runtime doubled to around 12hrs using £10 worth of extra battery. Or out to 24hrs if you have screen sleep enabled.

If/when you do need to stop and swap batteries, a Strava TCX/GPX file can be joined using one of a few methods to give that all-important one long ride.

The 4.4-star rated Viewranger app is free for Android and iOS. You can purchase Viewranger maps for less than half of the Garmin cost and the online route planning tool is genius. £90 gets you all of the UK (£199 from Garmin), or smaller bits are priced applicably less. Or you can download Openstreetmap / Opencyclemap tiles for free *from the app, on the mobile* and use them anywhere in the world. This can be done whilst on the road, no laptop needed - for example, using free Wifi in a café. Did this in Belgium last year- worked flawlessly.

Hopefully this goes some way to clearing the air and giving people hope that quality, robust on-bike long-ride nav, logging and comms is perfectly do-able for less than £100.

If you're new please join in and if you have questions pop them below and the forum regulars will answer as best we can.

Add new comment

88 comments

Avatar
KiwiMike | 9 years ago
0 likes

Road.CC reviewer in "Works well for your use case which should suit most people / budgets, other stuff is available" shocker  1

Avatar
SB76 | 9 years ago
0 likes

This'll do it:

http://support.apple.com/kb/HT1316

SIMLock, locks the phone from cellular data. You can then disable wifi BUT the GPS will still work.

Avatar
giobox | 9 years ago
0 likes

The raw GPS data that comes out of a phone is rarely a match for a dedicated device like a garmin in my experience. By necessity, mobile phones prioritise battery life before GPS accuracy, meaning it will often settle for good enough, rather than waste the energy getting a decent location fix. Having written some iPhobe apps that use apple's CoreLocation API (the means by which app developers access GPS on the iPhone), it's designed entirely to provide high battery life and "good enough" location fixes, rather than outright accuracy. I can't speak for how Android handles this, but I imagine a similar battery Saving strategy will be being used.

I've tried comparing GPX trails from an iPhone 5, samsung galaxy s5 and edge 800, and the edge works way better, especially in areas where the GPS signal might be poorer, such as narrow streets or tree covered lanes. Not a problem for navigation apps, but if you are trying to record accurate data to something like Strava worth bearing in mind.

Avatar
dave atkinson | 9 years ago
0 likes

I've been looking at the strata plot from this morning and I'm struggling to see why it would need to be more accurate. Nor am I convinced that a garmin would have been more accurate.

It's accurate enough that I can see the point where I pulled over to the side of the road to put my rain jacket on. It very rarely leaves the confines of the plotted road on the map.

Avatar
dave atkinson | 9 years ago
0 likes

Battery life not an issue on most rides, the z1 compact will manage at least 8 hours the way I use it, without changing any of the settings

Avatar
dave atkinson | 9 years ago
0 likes

Here's some screen shots overlaid on 1:25k OS mapping, showing:

1) out and back along the towpath, under tree cover
2) not coping quite so well on a short tunnel under the aqueduct
3) looping back to a petrol station to buy some midget gems
4) pulling over to put my coat on
5) general good roadholding throughout the plot

Avatar
SB76 | 9 years ago
0 likes

I would expect that a dedicated device would be more accurate but i havent done the testing however what level of accuracy do either cars/cyclists really need. I have used my old iphone as a car satnav and to do thta, i bought the tomtom cradle that came with an additional gps radio to boost the poor indoor/car quality gps of an iphone 3GS. I have some the odd strange behaviour using strava on my phone but where that comes from is difficult.

The key thing is Strava must perform some post processing as the stats for the rides do change between hitting save on the device and the ride being uploaded.

I believe the phone is suffcient from a GPS perspective so it's purely about battery life and other things such as waterproofing.

Avatar
KiwiMike | 9 years ago
0 likes

5 years ago I'd have been the first person to agree that stand-alone GPS was the way to go - as a former mobile network operator and manufacturer technology strategist, I have drawers full of now-obsolete GPS receivers, old phones, etc. Testing this stuff used to be my day job.

Now GPS has gone the way of bike lights. Used to be £200 was the entry point for lights good enough to do 50km/hr on unlit lanes for 2+ hrs. Now it's £20 or less.

As Dave / SB76 alluded, the conversation now moves to waterproofness and battery life - although increasingly battery clearly isn't an issue for the vast majority of rides, and the Quadlock silicone ponchos or case things like p0cpacs mean you can happily take a £500 non-IP-rated phone out on the bars or in a pocket in a thunderstorm and be happy.

I'd not be surprised if the next iPhone comes IP rated. It used to be IP rating a handset added $100 to the bill and quite a few mm to the dimensions. As with GPS, 8+ MP cameras etc, it's now almost a hygiene factor.

Avatar
Mr Will | 9 years ago
0 likes

I use an Edge 500 for my bike computer and for tracking and my phone (Xperia Z1 compact) for navigation and communications. Both are better at their respective jobs and I have no worries about my phone being dead should I have a problem on a long ride.

Why should it be a choice between one and the other?

Avatar
dave atkinson | 9 years ago
0 likes

Been playing about with mounting the phone on the stem over the past week or so. pics below.

the stem mount is just a standard o-ring one. the phone mount is a SRAM Quickview adaptor that allows you to mount an older garmin on a quarter turn mount, i dremelled it down so just the quarter turn bit was left

it's stuck onto a semi rigid skin with a 3M VHB pad, the same stuff GoPro use for their helmet mounts.

so far so good. the phone isn't in any danger of coming out of the skin which is a tight fit, but still has enough flex to give a bit of shock absorption. it also protects the back of the phone from hitting the stem, and means you don't have to stick a garmin mount directly on to your phone. I was going to get another skin for off-the-bike, but i haven't bothered yet, it doesn't stick out much.

Avatar
philtregear | 9 years ago
0 likes

i use the edge 200 for navigation. its great and easy peasy. i cant disagree that smartphones might be ok, i have no need or desire to wish to learn to use such tec. for £90 and out of the box ease of use the garmin is great. my cheap mobile sits in my pocket . both have enough charge for a full days riding and some left over if im out longer. also i could always use the torch on my phone to get me home!!

Avatar
parksey | 9 years ago
0 likes

That stem mount of yours Dave has convinced me that a Quad Lock is the way forward rather than the Finn, as that's exactly the solution I'm after. With a waterproof phone you don't need to worry about a separate "poncho" for it either.

Just a matter of finding a suitable case to stick the universal adaptor too, not much around for the Z2 yet.

Avatar
Judge dreadful | 8 years ago
0 likes

I only use the Garmin as a clock, when it's too dark to see the cheapo cycle computer screen. I use the smartphone with google maps set to navigate, for everything else, I have Stravabollix running to record the route. I regularly go over 12 hours riding, and have only had to use the mobile charger thingy on a handful of occasions. I've lost count of the number of times the Garmin has let me down, by being useless for navigation, and having a less than impressive battery life.

Avatar
SB76 replied to giobox | 9 years ago
0 likes
giobox wrote:

The raw GPS data that comes out of a phone is rarely a match for a dedicated device like a garmin in my experience. By necessity, mobile phones prioritise battery life before GPS accuracy, meaning it will often settle for good enough, rather than waste the energy getting a decent location fix. Having written some iPhobe apps that use apple's CoreLocation API (the means by which app developers access GPS on the iPhone), it's designed entirely to provide high battery life and "good enough" location fixes, rather than outright accuracy. I can't speak for how Android handles this, but I imagine a similar battery Saving strategy will be being used.

I've tried comparing GPX trails from an iPhone 5, samsung galaxy s5 and edge 800, and the edge works way better, especially in areas where the GPS signal might be poorer, such as narrow streets or tree covered lanes. Not a problem for navigation apps, but if you are trying to record accurate data to something like Strava worth bearing in mind.

i thought the iphone used aGPS which utilises Cell triangulation first then gps to fine tune. Got to admit, i have worked on GPS systems in the past but not for mobile phones

Avatar
giobox replied to dave atkinson | 9 years ago
0 likes
Dave Atkinson wrote:

Battery life not an issue on most rides, the z1 compact will manage at least 8 hours the way I use it, without changing any of the settings

I think people misunderstand what I'm trying to explain here. Of course you can write a GPS using app that will return good battery life. The point is that a compromise has to, and will be, struck between GPS accuracy and battery life. Techniques such as reducing location polling based on speed of movement etc wll be employed to maximise battery life on a phone. On an iPhone this can be a pretty big, although arguably necessary, compromise at times.

A device such as a Garmin does not have to worry about running background services for calls, texts, pinging cell towers, managing memory for third party applications, etc, etc, affording the engineers a degree of freedom that simply does not exist on a phone based platform. This is one reason why you can at times see some pretty erratic GPX data out of an iPhone, where the Garmin tracks more cleanly.

Of course, I haven't even touched on the iPhone's lack of an altimeter, but that's a whole other issue with location accuracy on phones.

Avatar
dave atkinson replied to SB76 | 9 years ago
0 likes
SB76 wrote:

I would expect that a dedicated device would be more accurate but i havent done the testing however what level of accuracy do either cars/cyclists really need.

i would expect a dedicated gps to be more accurate too, but my experience, and looking at plots off my edge 810 compared to my phone, suggests that they're not. or if they are, the phone plot is sufficiently good for it not to matter

Avatar
dave atkinson replied to Mr Will | 9 years ago
0 likes
Mr Will wrote:

I use an Edge 500 for my bike computer and for tracking and my phone (Xperia Z1 compact) for navigation and communications. Both are better at their respective jobs and I have no worries about my phone being dead should I have a problem on a long ride.

Why should it be a choice between one and the other?

it doesn't have to be. but if you're navigating on your phone then it's a sight easier to do that if it's on your bars.

Avatar
dave atkinson replied to giobox | 9 years ago
0 likes
giobox wrote:

I think people misunderstand what I'm trying to explain here. Of course you can write a GPS using app that will return good battery life. The point is that a compromise has to, and will be, struck between GPS accuracy and battery life. Techniques such as reducing location polling based on speed of movement etc wll be employed to maximise battery life on a phone. On an iPhone this can be a pretty big, although arguably necessary, compromise at times.

A device such as a Garmin does not have to worry about running background services for calls, texts, pinging cell towers, managing memory for third party applications, etc, etc, affording the engineers a degree of freedom that simply does not exist on a phone based platform. This is one reason why you can at times see some pretty erratic GPX data out of an iPhone, where the Garmin tracks more cleanly.

Of course, I haven't even touched on the iPhone's lack of an altimeter, but that's a whole other issue with location accuracy on phones.

I'm not misunderstanding you. I'm saying:

1) look at the GPS accuracy on the ride i posted earlier
2) my Xperia Z1 Compact is good for 8 hours on my bars, using the screen for navigation when necessary, at that GPS accuracy. at least 8 hours.

where's the compromise? I can't see one.

Not everyone's trying to do this on an iPhone, and if you do want to use your phone as a GPS then the iPhone would be a good way down my wishlist. not least because it's not waterproof.

whether barometric data is available is kind of moot when all the major platforms use topographical data on upload to determine altitude gain. that's a whole other conversation in itself, really.

Avatar
SB76 replied to dave atkinson | 9 years ago
0 likes
Dave Atkinson wrote:
SB76 wrote:

I would expect that a dedicated device would be more accurate but i havent done the testing however what level of accuracy do either cars/cyclists really need.

i would expect a dedicated gps to be more accurate too, but my experience, and looking at plots off my edge 810 compared to my phone, suggests that they're not. or if they are, the phone plot is sufficiently good for it not to matter

I think the key is sufficiently good enough not to matter is the point. We dont need the plots to be inch perfect.

Avatar
KiwiMike replied to giobox | 9 years ago
0 likes
giobox wrote:

The raw GPS data that comes out of a phone is rarely a match for a dedicated device like a garmin in my experience.

Hmmm....this PhD bloke in GPS/GIS systems has looked specifically at iPhones vs. dedicated GPS units, and begs to differ: http://blogs.esri.com/esri/gisedcom/2011/04/22/comparing-the-precision-a...

"This photograph ... turned out to be only 6.21 meters away (this time to the southwest) from my position as recorded by the GPS receiver. Interesting! This means that the SmartPhone was indeed recording an accurate position"

...and this was 3 years ago. Chipsets and antennae have come on in leaps and bounds since then.

Every comparison I've done between my iPhone 5 and Garmin 810 tracks has shown next to zero difference. In 7,000km of recording through Hampshire's often deep and winding, tree-shadowed lanes, I have never seen a single gap in the GPS track. And this is with the phone vertical in a pocket under 2 or three layers - not horizontal out front.

Mobiles and mapping apps are optimised for on-foot navigation. Footpaths are only a meter wide, at best. Using maps to navigate around London, you are shown on the correct side of the street. I use this every day to find client offices or meeting places. Apple / Samsung et al MUST deliver a location experience that is accurate to within a few meters, within a few seconds. WiFi is good for 100m, Cell triangulation for maybe 300-500m. But GPS is good for a few m. Most of the time, it works in cities. All of the time it works in the countryside (IMHO, waves 7,000km a year as proof).

Avatar
giobox replied to KiwiMike | 9 years ago
0 likes
KiwiMike wrote:
giobox wrote:

The raw GPS data that comes out of a phone is rarely a match for a dedicated device like a garmin in my experience.

Hmmm....this PhD bloke in GPS/GIS systems has looked specifically at iPhones vs. dedicated GPS units, and begs to differ: http://blogs.esri.com/esri/gisedcom/2011/04/22/comparing-the-precision-a...

"This photograph ... turned out to be only 6.21 meters away (this time to the southwest) from my position as recorded by the GPS receiver. Interesting! This means that the SmartPhone was indeed recording an accurate position"

...and this was 3 years ago. Chipsets and antennae have come on in leaps and bounds since then.

The example you cite is for a clear area in an urban environment - exactly the conditions that the CoreLocation framework works best in. I would expect a great match between a Garmin and an iPhone there, as the three data sources that CoreLocation taps will provide optimal information. My point is, when conditions get poor, the Garmin starts to outperform my iPhone significantly.

The issue is not wholely due to the chipsets, its really down to how you write your GPS using iPhone app, if you're curious you can read up on it here.. The API is designed to encourage the developer to prioritise battery life.

Long story short - you want Garmin level of accuracy over time on an iPhone, the battery life is gonna drop, rapidly. Hence the compromises in the quality of data most iPhone GPS logging apps return.

Avatar
KiwiMike replied to giobox | 9 years ago
0 likes
giobox wrote:

Long story short - you want Garmin level of accuracy over time on an iPhone, the battery life is gonna drop, rapidly. Hence the compromises in the quality of data most iPhone GPS logging apps return.

But I just don't see this supposed compromise - every ride since we went on Strava, my friend's 810 turned in basically the same distances and altitudes as my iPhone, more or less. It seems there's a natural 'noise floor' for both platforms, but over 50km they are usually within a few 100m or so of each other, with neither one showing strongly over a super-accurately-calibrated wheel-sensor speedo. Let's face it - Strava's own map info comes into play in a big way anyhow. I've compared Google Earth, MapMyRide, Google Maps, Strava, Viewranger and a few others - they *all* vary, be it in pre-planned estimates or post-ride track analysis. There is no golden reference, even on climbs in the alps which have been logged and mapped to death over decades.

Basically, pick a platform and stick with it for comparison with your efforts over time. Like comparing yourself on Strava against others who may be drafting/just starting, comparing GPS platforms for accuracy is a mug's game  3

Avatar
dave atkinson | 9 years ago
0 likes

Alors, so this morning i did this:

http://www.strava.com/activities/138781541/overview

which involved some inclement weather. very inclement. Here's what I found:

1) the xperia z1 compact really is waterproof. at least, it's still working after three hours in the rain (and mud) and being rinsed off in the shower

2) if you want to do stuff that requires high accuracy on the screen (texting etc) in the pissing rain then you may as well not bother, cause it doesn't really work. general swiping and fannying about is okay

3) i had a couple of rogue events on the screen where it flicked to some other view. but it was only a couple, in three hours.

I guess it depends on your use case. i have the phone on my bars and i turn on the screen occasionally to check i'm still on the blue line in Viewranger, and then i turn it off again. That means I don't ever have to stop, which makes rides more enjoyable and also shorter (a good thing when you have a family). I don't use it as a bike computer, and for my usage it works more or less perfectly. if you want to use it more like a standalone GPS (data constantly on, looking at lots of different screens, swapping between them, etc) or you're doing something like training with power where you'll need the screen on constantly then you're probably better off with a garmin (other gps units are available)

Avatar
dave atkinson | 9 years ago
0 likes

in related news, i'm chatting to the guys at raceware direct (the 3d printing chaps) about options for mounting a phone out front. more on that when there is more...

Avatar
dafyddp | 9 years ago
0 likes

I switched from using multiple apps on an iPhone5 to a really basic Garmin 200 - and much prefer the latter. I still carry the iPhone but given that the Garmin is only 58g, the weight penalty argument doesn't really wash. I prefer the Garmin for a number of reasons:

  • clarity - mono screen, works in all (day) light conditions, so no squinting to see where you're going (especially if your phone id in a case)
  • I like the simple breadcrumb trail - one you get the hang of it, it makes following a route really simple, and beeps if you overshoot or take a wrong turning
  • Low value and discreet (you can leave on your bike when you nip into a cafe)
  • simplicity of the interface - easy to switch between modes or pause even with gloves on
  • Avatar
    SB76 | 9 years ago
    0 likes

    And yes, it doesn't make the most sense! The rules are changing but they were for any comms at all causing interference.

    Avatar
    dave atkinson | 9 years ago
    0 likes

    https://strava.zendesk.com/entries/21292027-Extending-iPhone-Battery-Life

    GPS wouldn't be classed as dangerous by the CAA and FAA because it's receive-only

    Avatar
    SB76 replied to dave atkinson | 9 years ago
    0 likes
    Dave Atkinson wrote:

    https://strava.zendesk.com/entries/21292027-Extending-iPhone-Battery-Life

    GPS wouldn't be classed as dangerous by the CAA and FAA because it's receive-only

    It is intended to be disabled however:

    http://support.apple.com/kb/ht1355

    Avatar
    dave atkinson | 9 years ago
    0 likes

    you have to be careful using airplane mode on some phones (including iphone) - the GPS will continue to work once connected, but if it loses the connection it *won't* re-acquire it.

    Avatar
    SB76 replied to dave atkinson | 9 years ago
    0 likes
    Dave Atkinson wrote:

    you have to be careful using airplane mode on some phones (including iphone) - the GPS will continue to work once connected, but if it loses the connection it *won't* re-acquire it.

    Really????

    I find that difficult to believe but still highly believable, the whole point of airplane mode is to disable those parts of the phone that the CAA and FAA oddly believe to be dangerous to the systems of the aircraft. A GPS connection would most certainly fall into that category.

    I cannot believe that to be a design intent, more likely a bug - So dont shout about it...

    Pages

    Latest Comments