This is an expanded comment which I posted at here originally, but disappeared.
So, for the uninitiated the location based services field – and specifically the technologies surrounding the ability to find yours, or anothers, location – is a convoluted one.
I often find it both incredible and ridiculous that so much energy and time is expended trying to solve a problem, often for good reasons, which is clearly going be solved later on when the market, technology, or both, are ready.
Finding ones location (and all the functionality surrounding that) is a very good example. At Rummble, we “wasted” some considerable time early on, either looking at location or researching ways to solve it – including building our own DB of cellIDs etc. Thankfully, we didnt waste as much time as we could, and eventually excepted that we should stay focused on our core efforts in the knowledge (well, our own belief) that this would be solved well (and probably with access for free) by someone else.
Confirmation of this came in late 2007 when Google released its “My Location” function for mobile phones; a fantastically simple (atleast to the user, but not most likely technically) piece of software that shows on Google maps, on your phone, where you are.
Over the coming weeks I used in on my Blackberry 8800 (with GPS turned OFF) as it was often accurate enough, and quicker, than using the built in GPS. The GPS on the 8800 is great btw, it often locates in as little as 10 or 15 seconds – compare this to the N95 which at best takes 30 seconds but more often takes up to 2 minutes.
I’m told with the latest firmware the N95 has improved alot; but not enough. I gave up after the first two months – GPS was unseable slow and the overall battery life appalling. The N95 was the location-based services King-of-phones which never was. Its combination of functionality, size and specification (on paper atleast) should have made it a king to be admired in the anals of mobile phones for years to come – much like some of its predecessors, the Nokia 8310 or 6210; but it was not to be. Anyway, I digress, thats another post…
So, I’ve found myself 8/10 times using Googles MyLocation maps; its instant and has improved over time (as their database of cell towers improves and usage by hundreds of thousands of users enables the service to become more accurate). I was recently told by a Googler at a recent conference which is improving by 4% a week – that may have simply been an illustrative figure, but certainly using it I noticed this.
Last week I was walking down a main thoroughfair in London and could watch myself moving on the map as I walked – even HE was suprised to hear that. It was truly impressive – and again, that was without GPS turned on.
So LBS on phones are fast becoming a reality – it is the last peice of the jigsaw in my mind, to enable the mobile internet to take off in a proper mass market way. That still wont be until Q4 2009, but the table is laid and now we await the food!
There have been rummblings recently of Google releasing a MyLocation API. My personal opinion is that this is a very obvious one for Google to leverage their infrastructure to get a buy-in from the developer community.
Releasing this API but providing people/developers use GoogleGears – or more likely, Android (their mobile operating system) would add to the momentum of Android and getting developers aboard.
There are a handful of startup open projects trying to pull together Cell ID information, these include http://celldb.org/ and http://www.opencellid.org/ but with atleast 3 other location based startups I know trying to do this, I think the openDBs and the startups may be late to the party. I actually emailed both to suggest they work together and that I would help, but got no reply from either.
With Google (and others – Yahoo has a vast DB it could give access to in conjunction with Fireeagle, and the resources and reach to instigate a CellID effort) location, by GPS built into phones (I believe GPS will be the next “camera” in phones, i.e. the next ubiquitous standard piece of functionality) location is something that is both important but also fast being solved. Startups putting significant effort into solving this problem, are making a mistake in my view, as thankfully, its being solved by the big boys.
This time next year, if not before, there will be a geolocation service SDK/API available to all; and most likely, free.
The iphone has location built in, using Skyhooks in licensed form as I understand it, and a combination of cellid and wifi hotspots is used by their service to provide location.
Mobile IS the future of advertising – as Eric Schmidt was recently recording as stating, as if it wasnt obvious already; and for Google, with much riding on Android as its way to grasp an influencial foothold in the mobile landscape, they need to do everything they can to ensure its success. Making LBS easy (as it is on the iphone) is one obvious way to accelerate interest and takeup.
As an entrepreneur who has pitched more than one company to VCs, one is often faced with the inevitable question “And what if Google* decides to do this” (* replace Google with any analogous large blue chip organisation with much resource).
It is a fair question, but sometimes we do tend to over estimate the ability of large incumbants to react and innovate, or enter and then dominate new markets or sectors. Despite its prescence on mobile, I’d argue that thus far Google has not impressed on Mobile. It’s not as bad as Microsofts failure to embrace the internet early on; but I feel there is a “could do better” report card due.
This opinion is of course voiced without the insider knowledge of the Googleplex in Mountain View, or Googles London centre of mobile development. I am sure they have plenty more up their sleave than MyLocation. It is also fair to say, that Android could yet have huge impact. I have high hopes for Android, which is headed up by Google and a selection of other mobile industry stake holders. All I am saying is, that its going to be a much tougher fight for Android than people realise.
The mobile industry – or atleast, its traditional “owners”, are responsible for the delayed and continued painful transition from traditional mobile voice and text, to “the mobile internet”. The MNOs are directly responsible both for the more recent acceleration towards mobile internet adoption (e.g. Vodafone UK’s recent bundling of data on all its tarifs) and the false start WAP debacle in 5 years ago and the wall garden, niave “we want to be content providers” falacy, ever since.
One should not underestimate the MNO’s continued determination not to see their grip on cell phone users slip into the hands of others – including Google – and its this reason which has strangled location based services until now. Expensive, un-unified access to Cell tower information (MNOs still charge for lookups – 12 pence each on last time I had the conversation) and the handset manufacturers have failed (as usual) to provide a sensible reliable way to retrieve GPS information from the handsets (the J2ME GPS implimentation works but is buggy across handsets and I wasnt aware that it was possible via a mobile browser – although in Mike Butchers recent post these guys claim to do it).
So, in summary, obtaining a location on mobile if you want to automate it into a cross-handset application, is still rather convoluted, but its about to get much much easier. The biggest Google can do to the mobile internet industry is provide a comprehensive -and free- API for developers to leverage their MyLocation platform.
Whats important though, is that this is made available for all mainstream platforms, not simply GoogleGears on Windows Mobile, or only Android; given the advantages to Google of using the LBS as a hook to pull users in to using exclusively Google technologies, I suspect releasing it to work on non-related Google platforms, will not be a priority.
The other problem is one of maps. Many location based services want to use maps; currently Google maps Microsoft maps can be used on mobile – and this isnt their decision. Its dictated by the license agreements from the map providers. Currently the only solutions are paid for maps, from companies such as Navteq, now part of Nokia.
I’d argue that it will be one of the single most important contributions to the success of mobile applications, in the last 3 years.
Off the record recently*, I spoke with a Google employee who confirmed that an SDK/API will become available – but gave no time frame. It wasnt a suprise to hear this – its an obvious decision; but it was good to get confirmation; and certainly dilute any negativity I’ve recently had towards Google for other decisions, which perhaps were not as helpful as this one will be.
As soon as its available, Rummble will certainly be taking advantage of it.