IPTV, to OSGi or not to OSGi
Ok, it is over. I admit, I played hooky on the OSGi world congress for the last few hours to chat with Olivier Gruber, who is having some very intriguing ideas about how to reduce the Java footprint. This is the best part of the Java conferences; not so much the presentations, as well as these types of discussions. Especially during the business conference; you have serious time to use the hallways.
This does not mean I skipped all the sessions. The one I'd like to discuss was from Kumanan Yogaratnam and Eddie Drake of Espial and Myrio. They proposed the combination of OSGi and IPTV for settop boxes. Intrigued by their presentation I did some googling and saw the development of an exciting story. A story that made me see the beauty of IPTV married with the OSGi Service Platform.
IPTV is hot due to some interesting market mechanics. Fixed telephone lines, around us for more than a hundred years, are heavily threatened by the emergence of Voice over IP (VoIP). I know, my phone bill was reduced from several hundreds of euros a month to a mere 20 dollars (~17 euros). A year ago I switched to Lingo, which gave me free calls to the US, Canada, and Western Europe from France. Since then, when I call my chess mate next door the call is routed over the Internet, through the US, completely free. Skype, Googletalk, and others threaten to further eradicate the traditional income stream of the Telcos by providing free PC-PC calls.
So these very rich Telcos must adapt or die according to Darwin. The low hanging fruit is to provide Internet based streamed services because they already own the last mile of the Internet. Providing TV over IP and Voice over IP in addition to basic IP connectivity are no brainers; this set is also called triple play.
However, the question is: Will these rather simple services be efficient to save the Telcos? The margins on plain old telephony were embarrassingly high due to inertia of the end users and virtual monopolies. No other industry could have recovered so quickly from the hundreds of billions of losses they incurred with the 3GT licenses and overly optimistic acquisitions in the late 90s. VoIP is already being commoditized as my own American VoIP subscription demonstrates. Broadcasters (Terrestrial and satellite) as well as cable companies have been in the TV game for many years and will not roll over and play dead.
So what is the unique selling point that Telcos can bring to the table? I think there are two key issues where Telcos can succeed in generating new revenue streams that can not be emulated by the broadcasters and that are currently costly to emulate by cable operators. These are the exact interactive services that the OSGi has been talking about for such a long time. TV services like time shifting, availability in different rooms, MP3 players fed from a central media server, interaction with the residential infra structure, security services, using user location information (also outside the house), interactive games, and so much more. With a third party application model, the imagination is the limit; look at the profits generated by the 900 numbers, SMS based services, and the mobile telephony Java applications (MIDP).
Obviously broadcasters can not provide these types of services due to the one-way nature of their architecture. But why could cable operators not compete; they support IP? Well, they also have a vested interest in their infrastructure. Their existing settop boxes are quite, ehh, simplistic because they squeezed the last dollar out of it (and more). Their architecture is geared around providing modulated signals to the home; the IP traffic is just one of those signals. These guys are experts in content streams, interactivity is very foreign to them.
The brave ones inserted an MHP stack but did not dare to spent the money up front to give it enough flesh to enable a viable service market (The MHP crippled scope and technical problems did not help). Why should they spent money for a an unproven model? Before the Telcos went into panic mode, the cable operators had no competition. They received a good revenue stream from their streamed content alone. Unfortunately for them, their existing infra-structure will make it very expensive to provide a useful application model until their investments have been recovered.
In contrast, the Telcos are now in a unique position to capture the value added service market that OSGi has been touting for such a long time. The Telcos will have to invest in settop boxes to handle the IP traffic containing VoIP, TV, and Internet. Of course, they could make the same mistake as the cable operators and cripple their settop box to just provide the bare services; it will be sure penny wise. Or they could take a great leap forward and provide settop boxes with a third party application model and sufficient capacity to put the cable operators and broadcasters at arms length for the foreseeable future.
It will be very interesting to see if the Telcos have the foresight to embrace such a strategy or will accidentally leave this value added service market to future providers once the networks are fast enough, see my VoIP example. Enabling a new value added services market seems the only thing that can potentially generate the revenue streams desperately needed by the Telcos.
The OSGi Service Platform is obviously a very good foundation for such a standard; it was developed for this use case. Not only does it play very well with existing standards like MHP and MIDP, it also enables a market of middleware. OSGi based middleware simplifies complex applications, thereby reducing the entry cost for new types of applications.
The OSGi Alliance therefore more than welcomes any parties that wish to develop a quality profile for the IPTV settop box. We could create this profile in a short time because the architecture, components, implementations and developer tools are proven and well in place as demonstrated by the vehicle and mobile markets.
If you need my help to make this happen, please contact me. Personally, I would really miss these good old familiar Telcos if they followed the fate of the dinosaurs ...
posted by Peter @ Monday, October 17, 2005