I'm looking into the new background location service options in the iPhone 4 SDK. It allows an app to run in the background and receive location updates from the device.
There are two methods offered. One is a battery intensive mode that continuously gets location updates. The second recommended method sends the app location updates when there has been a "significant location change".
Does anyone know what a significant location change might be? Is a 30 foot walk considered significant, or is a 10 block walk considered significant? I imagine it also depends on the accuracy of the location mechanism being used at the time.
I've recently done some field testing of the new background location service to get an idea of what constitutes a significant location update, what kind of accuracy to expect for the location hits and our general experiences using it.
The results are detailed in a fairly lengthy blog post:
iPhone Background GPS: Accurate to 500 meters, not enough for foot traffic
As Steve Jobs mentioned in the OS 4 introduction, the low power mode uses cell tower triangulation and does not activate GPS unit. Since the iPhone phone module needs to keep a connection to the cell network anyway, there should be no impact on battery life.
Since the precision of a location fix with cell tower triangulation is anywhere between a few dozen meters (in dense city locations) and a few miles, I think 30 ft is not a significant location change. I don't know the specifics, though (and as mentioned by the commenters, the Apple dev forums are the right place to talk about those).
Related
I am trying to set up an app that will be able to check people's locations in the background, see if they are at a given location, and send a ping to a server if they are. We do not want to drain the energy of our users, so we are attempting to figure out the best solution.
I've done substantial reading and I have not found very much information on these methods. I'll go through the pros and cons as I understand them right now
startMonitoringForSignificantChanges
Description: Based off of wi-fi and cell tower changes the system wakes up the app.
Docs:
Apps can expect a notification as soon as the device moves 500 meters
or more from its previous notification. It should not expect
notifications more frequently than once every five minutes. If the
device is able to retrieve data from the network, the location manager
is much more likely to deliver notifications in a timely manner.
Pros:
Most battery efficient
Cons:
Dependent on wi-fi/cell tower changes
Can only assume that this will be called every 200m to 2km (if not more in certain areas)
More on accuracy
Thus, inconsistent and imprecise
10-minute start-updating or "n-minute updating":
Description: This basically asks the app for more time, when that extra time is about to expire, it calls [self.locationManager startUpdating], grabs the location and extends the background thread for 10 more minutes.
Pros:
Consistent
Can be as accurate as you want it to be as consistently as you
want it
Cons:
Has to do a call every ten minutes or less to keep the app running in the
background (ie n can't be greater than 10 for the calls)
Questions:
What effect does this have on the battery? Does waking up the GPS and shutting it off hurt the battery more? I couldn't imagine running a brief location check in the background would drain the battery that much... but then again, I don't know what goes into powering up the GPS and getting a usable signal.
startMonitoringForRegion (geo-fencing):
Simply put, your app gets woken up when you enter into a pre-defined region. This is the oddball of them, it is more recent and there is less documentation on it. I can't find a good description on how the "system monitors" the boundary crossing. For all I know it is some really smart algorithm, or they are constantly pinging the GPS which would make it less effective than the other methods for doing this.
Pros:
Simple implementation
Managed by the system so you don't have to invent your own ad hoc geo-fences Only triggers on boundary crossing... no unnecessary data to just throw out in exchange for a battery hit
Thus, should be the best for this sort of thing, accurate, managed by the system
Cons:
People question its effectiveness
Huge conflicts on whether or not it is good for battery life or if it
drains battery life terribly.
How is the system monitoring this!?
Basically, indeterminate behavior.
I guess my question boils down to how does startMonitoringForRegion: compare to these other methods of testing user location in the background when it comes to battery life, consistency, and precision. Has anyone thoroughly tested this? Or used it in their app and gotten at least some feedback? Likely, for my purposes, the trade-off is between geo-fencing and the 10 minute update method. (Also given what Apple has publicly said about iOS7 there will be some background tasks... will this change the calculus for the trade-off between these two methods?) Does anyone have an idea of how these two compare?
Thanks so much! Looking forward to seeing if we can get to the bottom of how to compare these methods.
I've been working on vehicle tracking using GPS for 2 years. Learned a lot the hard way... In my experience startMonitoringForRegion or Geo-fencing depends upon cell change events, didEnter or didExit events doesn't fire up until there is a cell/wifi change event. So it doesn't make any difference w.r.t battery consumption. However it does extra computation which depends on how many regions currently are being monitored. Even Apple's Reminder app doesn't give good results for location based reminders because it uses geo-fencing.
The other approach starting GPS for n minutes after each m-minutes is good option, it should not affect the battery life, if done wisely. What exactly effect the battery is constant GPS activation in high precision mode. Like for instance If you enable GPS with kCLLocationAccuracyBest and distance-filter = 0, you can literally observe battery drainage and soon your device will also start getting hotter.
If I was you, I would go for activating GPS after every 10 minutes for 5 sec with kCLLocationAccuracyBest (or may kCLLocationAccuracyNearestTenMeters to use less battery, if accuracy is not that much important) and distance-filter = 5 (meters). Battery consumption in this case will be unnoticeable. You can play around with similar settings which can address your specific case and finally find out what is best you.
BTW: iPhone uses AGPS, A-GPS additionally uses network resources to locate and use the satellites in poor signal conditions. So, when you do startUpdatingLocation it will also use nearby cell tower information. see http://en.wikipedia.org/wiki/Assisted_GPS
What I want to do:
I want to have my iPhone to frequently (formally defined later) upload my GPS location to a central server. I want to do this in the most battery efficient way.
Research I'm aware of:
Apple Documentation:
https://developer.apple.com/library/ios/#documentation/CoreLocation/Reference/CLLocation_Class/CLLocation/CLLocation.html#//apple_ref/doc/uid/TP40007126
Stack Overflow Links:
Response 1
Response: proof of existence; some other tool can do it
How to reduce iPhone battery consumption while using GPS
Response 2
Response: track only cell tower changes
iPhone GPS - Battery Draining Extremely Fast
iPhone running periodical process in the background - battery optimized way
Question
My question is a bit vague in that my definition of "frequently" is really dependent on what the battery life can tolerate. For example, if the battery can take updates of every 5 minutes, I'd like to do it every five minutes; if the battery could do this every 10 seconds, I'd like to do it every 10 seconds.
I really want to understand
the different ways (change on significant location, timer, background?) continuously uploading GPS locations can be implemented
advantages / disadvantages
approximately how long the battery life can last in each case
This seems like a fairly generic and common problem. Does anyone know of either:
an in depth analysis of the various methods
or if there is a single "optimal" way to do this?
[Moderators: feel free to mark this comment wiki. I'd love to just get lots of different answers + cost benefit analysis of them.]
Have a look at the Apple Documentation Here.
The significant-change location service offers a low-power location service for devices with cellular radios. This service is available only in iOS 4.0 and later and can also wake up an application that is suspended or not running.
It all depends on the use case. If the user will stay at a certain location and all you want to do is to track if he is leaving the country, tracking the significant location changing will be most suitable. However, this will not be precise enough for navigation apps.
One advantage of choosing the "Apple-Algorithm" is that they'll optimize it for you, if battery issues occur. (-:
I've heard of geofencing and I have some doubts. Hope you can answer some of them.
-It's new for iOS 5.1 but what is the difference between geofencing and Using Regions? From what I've read is nearly the same, the battery drain?
-Why can't I find apple documentation on geofencing?
-How do you define the regions? can I add them remotely, via webservice?
Thanks.
Geofencing has actually been around since iOS 4. You can find documentation on it in the Apple Location Awareness Programming Guide. It is also referred to as "region monitoring".
The regions are pretty simple to use. Create a CLRegion with the center point (lat, long), radius, and a unique identifier. Then call the -startMonitoringForRegion. Granted you will need to register a CLLocationManagerDelegate to listen for the callbacks, -didEnterRegion and -didExitRegion, but that is pretty easy to do as well.
You can't add them remotely. Only from the device. I have learned a few things about them over the last couple years. So hopefully this will save you some trouble down the road.
Not all devices can monitor regions, so make sure you check for availability
There is a min and max region size. If you register a region over or under those values, it will default to the min/max. The min is 100M, max is something like 1/3 the size of the globe. Pretty large.
There is a max number of regions 1 device can monitor. This "pool" of regions is shared by all apps, and older fences will get purged to make room for newer fences. So be prepared for that. The max number is close to 100. But shared between passbook apps, reminders, and other apps... can get burned up pretty quick.
Make sure you only have one instance of your location manager delegate, multiple instances will get you multiple delegate calls, they are all listening.
Geofences are a great way to add functionality to locations without any of the overhead active GPS and location tracking has. In my experience, geofences pose no noticeable battery drain when leaving on all the time. Apple has them figured out pretty well. They don't carry the same accuracy as active GPS, but they are close enough. Good luck.
UPDATE
Some of the information here is a bit dated now. The number of regions you can monitor is now 20 and is capped per app. Also, the minimum size is not 100M, it can be smaller, but accuracy is not guaranteed at lower numbers.
"Geofencing" has always been around, since iOS 4 in the form of "Region monitoring". It uses distance checking to see if you've entered a circle. This is not true geofencing, as true geofencing is checking if a point lands in a polygon (called point in polygon).
As of iOS 6, there is currently nothing akin to region monitoring that uses point in polygon. To do true geofencing, you would have to use a combination of Region monitoring and GPS. There is no Apple documentation for geofencing because CoreLocation doesn't support it.
If you are going to take the point-in-poly approach, I would recommend ray casting. There is an Objective-C class that implements it really well. It's called objc-BorderPatrol.
As for sending fences from a web service, it may be possible to do something similar to Twitter's streaming API, but that may be too intensive. It's better to use Significant Location to update your fences. If you're going to send fences from a web service, I would recommend using MySQL or PostGRE SQL as both have really good geospatial extensions.
A few caveats:
-> -[CLRegion containsCoordinate:] is never called by any CoreLocation object. All of the distance checking is internal.
-> Significant Location updates every 500m or 5 minutes, whichever comes first. Significant location uses the IP address from a cell tower to determine location.
-> Geofences and regions can overlap, keep track of this.
I would like to share the info of my performed testing scenarios and to ask you to share your experience with region monitoring. So, I have registered two opposite regions (A and B) with 5 km radius and 1 km desiredAccuracy, with the 20 km distance between them. However, I haven't received any didEnterRegion/didExitRegion events when traveling from A to B. In my app prototype, I also put two buttons (for testing purposes): one starts significant change and another standard location monitoring. Both prints didUpdateToLocation events on-screen-log.
So, after getting to the center of region B and being very annoyed about the unpredictable functionality, I have enabled significant monitoring. After that, I have received several didUpdateToLocation with horizontalAccuracy of ~8.5 km! Wow, that hurts, because all I want to do is to have a functionality where a user would be notified, that his desired place is somewhere in radius of 5 km from his current location position, and I want to rely on GSM cell triangular location detection method only (for preserving the battery).
So, back to my case. After getting such a big 8 km error, I have enabled standard monitoring with best accuracy (GPS). And bingo, I have received didUpdateToLocation event with horizontal accuracy of 399 m, and only then, I have received didExitRegion from my initial place (region A), and another event with didEnterRegion (B). Another updates from didUpdateToLocation where a little bit more accurate - 50m.
In previous scenarios, I was getting unstable accuracy also. Sometimes there were 4 km horizontalAccuracy, sometimes less. But the main question that is, if we want to preserve the battery (Apple teaches us that in docs) then can we rely only on region monitoring + GSM without using GPS? There's a sample of Apple code with "proper use of region monitoring" (here), however there are some unacceptable things there:
1) They state that sample code works only for iPhone4, though Apple docs state that both region monitoring and significant monitoring are supported on iOS4.0+. Cool, ha?
2) The sample code uses all 3 location methods: significant, standard, region. It uses region monitoring all the time but in addition in uses standard location with best accuracy. When app enters background it stops standard and starts significant change monitoring. When entering foreground then app stops significant and starts standard location. Seems like apple docs are quite vague, it's unclear what penalties (horizontalAccuracy) do we get if we use only region monitoring, or do we have always to use all three methods?
3) Apple docs state that we need to preserve the battery and to use as minimum location services as we need, if we don't need GPS then we should use GSM (for example, 1 - 3 km desired accuracy). But the sample code uses best desired accuracy (GPS) both for region monitoring and standard monitoring (when app is in foreground). I do understand that GPS would solve all my problems with errors and accuracy, but what about using GSM only? After getting 8.5 km horizontalAccuracy in my scenario, I'm not sure is it possible to use region monitoring at all because it's unable to provide at least 5 km acceptable error.
In my experience, the region monitoring is pretty accurate without a significant loss to battery life. I have a use scenario of regions that are only 30-50 meters wide. No issues at all with accuracy. One thing I will say, I have only targeted iPhone 4 devices and up. I will need to implement some of the changes you talk about if I want to support the 3GS models.
When relying purely on -startMonitoringForRegion, your updates automatically trigger on the -didEnter and -didExit events. These are triggered through a combination of location tracking events. Significant changes, cell tower transfers, connection to WiFi, another app requests location, and a few others. The OS handles when the callbacks are triggered from these other shared location events. In my experience, it has been very accurate. But this is only for iPhone 4 and up.
I also used the example app from Apple to get things set up, but there isn't a lot of detail in the documentation about the specifics of the when/where you will get your trigger. You can read up on what Apple provides in the Location Awareness Guide. Hope this helps.
quick question. How accurate is the GPS on the iPhone 4? I ask because I'm working on an enterprise project for a company, and part 2 of it will deal with iDevice development where I have to determine the position of the user. I'd like to know if the GPS is accurate enough to sense the user moving within rooms because the user will have to "tag" sections of the room as they move about it.
Thanks in advance!
P.S. Pressuming that it won't make much of a difference, but the users will actually end up using the iPads, not iPhones, and more than likely the iPad 2 will be out by the time the entire project is completed. I don't know if the iPad 2 will have a better GPS receiver or not, but at the minimum I should use the iPad/iPhone 4 GPS receiver...
Most buildings will not allow reception of an accurate set of GPS signals (if they can be received at all) indoors. The roof/ceiling/floors above are just too thick. Even a lot of trees overhanging a building will degrade the signal from the GPS satellites.
You might have a chance if all the rooms have very large unobstructed windows with no overhangs, and it's the right time of day for several satellites to be in view out that window.
Outdoors, in the clear, the iPhone 4 GPS seems to be very accurate. Sometimes I can walk around my parked car, and see the blue dot in the Maps app follow me in a circle.
I have done some work with a large location data set. My result set is based on cars driving outside and will therefore be, on average, more accurate than those taken inside (based on line of sight to satellites).
For the 650,704 location updates I used in my tests, I found the average accuracy radius was 246m (91m if your remove >1km outliers). 85.1% of updates had an accuracy of less than 100m. So given that your update will not be as accurate as these, I don't imagine you will have much success tracking indoor location changes.
For a further description of my results.
It is very difficult, and most of the time impossible to obtain a GPS signal inside a building. The type of waves used by the GPS (radio waves) are not powerful enough to go through the structure itself.
A simpler and probably cheaper solution would be to give people maybe tags or cards and install some sort of trnasreceiver in each room.
It seems the original question was "how accurate is the GPS on an iPhone 4", which hasn't exactly been answered yet.
I've done lots of testing with the accuracy of the GPS chips in iPhone 4, iPhone 4s, and iPhone 5, and the most accurate reading allowed seems to be ~5 meters, or ~16 feet when you're outside with clear line of sight to the sky. I'm guessing this is a software limitation imposed by Apple to conserve battery.