Telco Talk

In Costa Rica, the government owned telco - ICE is being forced to open up it's market to competitors because of the Central American Free Trade Agreement (CAFTA) that Costa Rica has joined. This represented a huge change for ICE who were a Power and Communications provider, without a competitor in their market, they didn't have any competitive forces to push them to modernise their systems and processes. For instance, fulfilment of basic services took weeks as a result.
 
GBM, an IBM business partner and IBM Software group proposed to ICE that they base their new OSS/BSS architecture on the TeleManagement Forum's Frameworx (eTOM, TAM, SID, TNA) - for which they used the WebSphere Telecom Content Pack and IBM Dynamic Process Edition to ensure ICE would have the standards compliance and dynamic BPM capabilities. By using WTCP and DPE, ICE reduced the effort required to build and deploy their new processes by an estimated 20-50%. A fundamental principle of Dynamic BPM is the Business Services layer which sits on top of the BPM layer which in turn sits on the SOA layer. A Business Service is abstracted up from the physical process. For instance, a business service might be 'Check Technical Availability' which would apply regardless of the service you are talking about - mobile, POTS or xDSL. These business services are defined within the Telecom Content Pack which enables system integrators like GBM to accelerate the architecture work on projects like this one for ICE.
 
GBM made use of IBM's Rapid Delivery Environment (RDE) - where they sent a number of their architects to the IBM Telecom Solution Lab in Austin, Texas for six weeks to conduct a proof of concept and to learn how to apply WTCP to a real customer situation such as that faced by ICE. The RDE allowed GBM to work with the IBM experts to build the first few scenarios so that GBM could continue the work locally in Costa Rica without a lot of assistance from IBM. The other benefit of using the RDE is to get access to the eTOM level 4,5 and 6 assets - the connections to the physical systems that the RDE has previously developed. For instance, the connection to Oracle Infranet Billing engine which can then be reused by other customers who also engage with the RDE.
 
GBM and ICE have not yet been able to measure that acceleration that WTCP and DPE provided, but anecdotal evidence suggests that it was significant. In preparation for CAFTA, ICE have already launched a 3G network and are preparing to launch pre-paid services in preparation to compete with several new operators that will enter the market this year.
#ibmimpact

 At this morning's keynote session with Beth Smith (IBM) and Shanker Ramamurthy (IBM) one of my customers - Globe Telecom from the Philippines was mentioned - unfortunately they could not be here to see if for themselves, so I thought I would post the photos and short video I took of it. <edit> I have replaced my shonky video with an extract of the relevant section form the official Impact videos on youtube </edit>.
 
 


 


Some official videos have been uploaded to YouTube - taken from the real (read good quality) cameras at the event. I have extracted out the relevant Telco section and added subtitles to clarify what Beth and Shankar are actually saying.
#ibmimpact

  AT&T are part way through a major SOA/BPM project which if you know a little about their history* must be an enormous task. They are introducing modelling tools and reverse modelling their existing systems as well as using a tool from iRise to prototype the user interfaces and reduce the risk of not hitting the business requirements.
 
They have deployed Rational Requisite Pro to capture requirements without the need to get users away from their beloved MS Word. In the last five months, their requirements have gone from 15,000 requirements registered in January to over 30,000 now. Certainly illustrates the traction that they are achieving with their business people. Users access Req Pro via Citrix sessions and the tools are available to thousands of business users.
 
AT&T are also exposing WebSphere Business Modeler and iRise to a smaller set of subject matter expert users - building a Centre of Excellence in UI design and Process Modelling. So far, they have modelled over 800 process flows base on eTOM models which have been extended to meet their specific requirements. All of these are stored within a common Rational Asset Manager instance which helps their business analysts to improve asst use and reuse.
 
Those process models feed directly into the model driven development method which is aligned with the requirements and process models. That MDD method uses WebSphere Integration Developer(WID), Rational Software Architect (RSA) for development and WebSphere Process Server (WPS) runtime. WebShere Business Modeler and WebShere Services Registry and Repository (WSRR) in support of the runtime. IBM GBS have put in place processes to support AT&T's development life cycle and governance requirements.
  
Key success factors that AT&T see include:

• Solve Critical Business Problems
• Win over senior Exec support
• Achieve Business Partner Alignment
• Integrated Tools Approach
• Organisational transformation
• Infrastructure investment
• Communicate, communicate, communicate!

 
 

---

* AT&T have been through multiple de-mergers and mergers and acquisitions over the past 10 years resulting in a hugely complex IT environment.
#ibmimpact

I have just seen Amy Wohl of Amy D Wohl Opinions present on Cloud computing, she was going through the various cloud models and spoke about Community Clouds. What she means by that is multiple community focused clouds as part of a larger (private) cloud. An example of that is the Vietnam Government that bought an IBM Cloudburst to provide multiple virtual private clouds to small businesses in Vietnam so that they can have access to computing power that they otherwise now be able to afford. For Telcos, this could be an offering to their local community groups - perhaps a local schools, bar, sporting clubs, service clubs etc but also potentially for commercial organisations - perhaps to small businesses.
 
She also made the interesting point that (in her opinion) we are too early in the cloud evolution to actually define standards. She believes that any standards set now would stifle innovation in cloud technology and interoperability. I was interested to hear about this since I attended a web conference call a few weeks ago run by the TeleManagement Forum's effort to create standards around clouds, particularly For Enterprise use rather than public clouds. I guess the Enterprise cloud market is the most likely type of cloud user that will need interoperability first, thus the emphasis on standards.
 
Amy co-presented with John Falkl from IBM who discussed BPM within the cloud. Given BPM is a business function, items subjects such as Security are usually one of the biggest hurdles for Cloud Services. There are multiple factors that fall under the title of 'security' such as encryption, roles, authentication (especially when using federates or external authentication services), legal data protection requirements and authorisations. John also pointed out a number of considerations that should be considered in enterprise cloud services including Governance models (which he sees as an extension to normal enterprise governance models). John's view of standards for Cloud services is that it will most likely start with Web Services standards such as WS-Provisioning and mentioned that there were multiple efforts around cloud standards. I might see if I can have a chat to both John and Amy after the session to get their views on the TMF's efforts around cloud standards. If that discussion is interesting, I will report back.
 
Amy made a really interesting point during the Q&A - she said that when she was at Microsoft a few weeks ago and asked about transactional activity in their cloud - they said that MS could not do it.... Very interesting especially when you consider that transactional integrity is a core capability on IBM's cloud capability.
 
<edit>
I asked Amy about the TMF Cloud standardisation - she hadn't heard about it, but did say that she thought that TMF's approach was right - asking the enterprise customers to specify their requirements - she also thought they were probably the right place to start for any cloud standards too.
</edit>
#ibmimpact

Gridit is a Finnish company that is providing online retail services which was only founded in 2009. They are owned by nine local network providers. Think of them as an aggregated application store that sells a broad range of services and products from those nine network companies as well as third party content providers. They plan to sell services and content such as:

• Music
• Games
• e-books
• Access
• Data storage
• Information Security
• Home services
• VoIP
• IPTV

 
They do not make exclusive agreements with the content/service providers and provide their customers with freedom of choice. For Gridit, the customer is king - they will seek out new content providers if there is demand from the customers. Gridit also interact with local network providers and 3rd party content providers giving the customers a single point of contact and billing for the services that they resell.
 
What Gridit are providing is pretty similar to an app store solution we deployed last year in Vietnam which was also a joint venture by a number of Telcos and a bank which provided a retail online store for products and services from those communications providers as well as 3rd party content providers except that Gridit are also offering a hosted wholesale service - I could go to Gridit and build my new company 'Larmourcom' and offer products and services from a range of providers that Gridit front end for Larmourcom. Gridit can stand up an online commerce portal for Larmourcom and also provide an interface to the back end providers to allow for traditional and non-traditional service assurance, fulfilment and billing processes.
 
To achieve this abstraction from the back end providers, Gridit have used WebSphere Telecom Content Pack to provide an architectural framework and accelerator for all of those services. IBM has helped Gridit to map those processes as defined within the TeleManagement Forum's standards (eTOM, TAM, SID) and map them to the lower level processes to wherever the content or services come from.
 
Like the Vietnamese app store, Gridit are also using WebSphere Commerce to provide the online commerce and catalogue. For Gridit, the benefits they expect to see (as a result of a Business Value Assessment that was conducted) was 48% faster time to value by using Dynamic BPM and Telecom Content Pack versus a traditional BPM model. That is real business value and a great story for both Gridit and IBM.
#ibmimpact

Orange in France are using WebSphere sMash to provide an easy development environment using PHP and Groovy to build Telco enabled applications that consume Orange Application Programming Interface (API) which are exposed through pre-built widgets. The custom Orange API is not compliant with either OneAPI or ParlayX and I would normally not endorse a custom API like this, but time to market forces meant that Orange had to move before the (OneAPI) standards were in place. What I would take from their experience in France is their model and use cases. All of which could be done and (now) use standards for those APIs. Interestingly, I think that Orange could also use IBM Mashup Center to support developers with even less skills that the PHP and Groovy developers they're currently targeting.
 
#ibmimpact
Once I get back to my PC, I will insert an Orange video the positions the usage and simplicity of their offering.
 

Telus is a Communications Service Provider in Canada, the second largest in their market with 12M connections (wireline, mobile and broadband). Telus have a very complex mix of products, services and systems and they need to maximise their investments while still be able to grow and maintain a lid on their costs. New projects still need to be implemented through good times and bad, so they need an architecture that will allow Telus to continue to grow and maintain costs through a range of economic conditions. Telus selected an agile method/strategy where a reasonable investment early on with the plan to become agile and support new 'projects' through small add ons in terms of investment. Ed Jung from Telus characterised the 'projects' in the later stages as rule or policy changes which may or may not require a formal release.
 
To achieve this agility, Telus are using WebShere Telecom Content Pack (WTCP) as an accelerator to keep costs down, while still maintaining standards compliance for their architecture. He sees key success factors as:

• Selecting a key implementation partner (IBM)
• Using standards where possible to maintain consistency

 
For Telus, they elected to start with fulfilment scenarios within their IPTV system. The basis for this is a data mapping to and from a common model - within the TeleManagement Forum's standards, that relates to the SID. Ed sees this common model as key to their success.
 
Dynamic endpoint selection is used within Telus to enable their processes to integrate and participate with their BPM layer. Ed suggest the key factors for a successful WTCP project are:

• Adopt a reference architecture
• Select a good partner
• Seed money for lab trials
• Refine architecture
• Choose correct pilots
• Put governance in place (business and architects)
• Configure data / reduce code

 
Ed thinks that last point (configure data / reduce code) is the best description of an agile architecture that really drive lower total cost of ownership for projects as well as a lower capital expenditure for each project.
 
 
#ibmimpact