Cable IT: JacobsRimell

Back to January 2006 Issue Who’d have thought when the Kinks sang about "a mixed up, muddled up, shook up world" in the 1970s that they’d be describing today’s telecommunications space? What better explanation is there for how JacobsRimell, an operations support system (OSS) vendor that started in the IT department at Reuters eight years ago, became today the service fulfillment provider of choice for cable’s biggest player, Comcast? JR, after all, isn’t a cable company, in the old-boys-network sense. Founders David Jacobs (now CTO) and Phil Rimell (now chief architect) designed Reuters’ information architecture and all the major components in its voice and data network backbone. That was, of course, not quite like designing the first electro-mechanical set-top box. But it led to cable and, eventually, Comcast. How? The accidental twist "I think it’s more probably by accident than design," says Joe Frost, JR’s marketing vice president. The company’s first deal was to provide dial-up throughput and resource allocation service for Cable & Wireless. That morphed into TV provisioning for C&W and NTL, the cable company that acquired the carrier. That then led indirectly to winning the AT&T Broadband provisioning infrastructure contract, which, all things being equal, would seem to have opened the door to Comcast. But all things are not equal. Upon acquiring AT&T, Comcast cancelled the JR contract. "We had to re-bid and won it again," says Frost. Meanwhile, thanks to what many call the telecommunications nuclear winter earlier this decade, cable was a lot safer business than anything to do with the imploding telcos, so JR concentrated there. It’s only been in the last couple months, in fact, that the company has begun pursuing what would seem to be logical customers for OSS provisioning: telephone carriers. Not de-focusing "We’re not in any way de-focusing our resources or efforts on the MSO space; we’re building the team to add resources focusing on the telco space as well," Frost asserts. Whatever the customer target, JR has, almost from the start, kept its focus on Internet protocol (IP) and the converged services that move over the IP layer. "Most of our activity has been on voice-over-IP (VoIP)," says Frost. The company sharpened its VoIP focus to help operators offering VoIP with home-grown, in-house provisioning platforms. That method, Frost says, doesn’t work; it doesn’t scale, and it takes control from customers who would like to select and configure their voice service with minimal operator interference. What’s exciting "Our challenge, what’s exciting for us, is to get in there to solve an existing problem and show the customers how they can start to converge their entire operation starting with the information management aspect first," says Frost. In a way, although Jacobs Rimell’s route to cable success seems serendipitous, Frost says a theme has wended through the company’s history starting with Reuters when Jacobs and Rimell decided "the operator … has to think about the customer’s needs first rather than think about it from a technology point of view. That’s where the subscriber-centric architecture which makes up our core solution set came from." Here follows an interview Q&A with JR VP Joe Frost on the sub-centric architecture: Could you outline the Java 2 Enterprise Edition (J2EE) structure of your Adaptive Provisioning System (APS)? What else can you say about the software you employ? Being J2EE-based, APS is a Java application by definition. It’s written to ensure it doesn’t make use of any individual platform vendor’s special extensions—so it runs on any application server infrastructure. The system is highly componentized and uses extensible markup language (XML) documents to communicate between components. With extensive use of Java architecture for XML binding (JAXB), this ensures better system performance in terms of concurrency of transactions and responsiveness. A key requirement of a multiservice fulfillment platform is information "openness" and performance, when you consider how many subscribers (tens of millions) and how many support, service delivery systems and infrastructure elements must be configured and managed in real time. How does your real-time database solution interface with an internal cable operator’s database? Is this outside the cable operator’s network or part of it? In order to maximize customer service effectiveness, right first time service fulfillment, and the ability to provide at least what looks to the customers like a converged operation, all the systems need to refer to a common information model—one that is federated and synchronized with a normalized copy of the relevant data. To achieve this, our preferred communications with all systems is via simple object access protocol (SOAP) or XML-based protocols—integration with databases can make use of these mechanisms or via direct use of Java database connectivity (JDBC) for relational database management system (RDBMS) and Java naming and directory interface (JNDI) for lightweight directory access protocol (LDAP) based directories. In terms of whether it’s inside or outside the cable operator’s network, the answer is "it depends." Ideally it is, or is a part of the core information architecture; however, depending on the stage of evolution toward IP management subsystem (IMS) and/or convergence, the data model may sit alongside the existing systems and provide some of the synchronization and integration functions in the early stages. How would you characterize billing system infrastructures that you have encountered in the cable industry? Are you assuming roles previously held by those providers, or are you fulfilling what are arguably new functions? The trend is generally toward decomposition of billing systems infrastructure (across cable and telco operators), and on a case-by-case basis, provisioning systems are taking on some of the roles previously managed by the billing system vendor because billing systems used to do many more functions on top of billing and are now being redirected back to delivering billing functions. Decomposition allows the operator to have more choice and flexibility. One key area under discussion still is that of the product catalog—in our opinion, this area should be split across the business support system (BSS) (billing and customer relationship management) and the OSS—each plays a part in implementing this successfully. Do you see your work more connected to a cable operator’s IT or engineering department? It depends. If a service relies on provisioning of quality of service (QoS) and down to the customer premises equipment (CPE), such as a set-top box, then it’s both IT and engineering. If it’s purely delivered over an existing infrastructure without QoS, then IT only. Richer services need end-to-end delivery infrastructure visibility and guaranteed QoS—this then involves both IT and engineering organizations. Do cable operators need someone in a CIO position? Yes—information engineering is a key element to drive effective customer centric and converged operations. If you believe a cable operator’s operation needs to trend toward customer centric operations—with all services tied to the user’s identify and preferences (particularly as we move to a more IMS-oriented model), then CIO is a key role. Additionally, as the operator starts to converge operations and services, the CIO and CTO need to be joined at the hip! How and with what protocols does your "subscriber-centric" technology interface with other network elements in the cable plant? Preferred information exchange is SOAP or XML-based; however, in practice, we have to interoperate with all systems of all ages, so we have to use pretty much every protocol out there today, ranging from common object request broker architecture (CORBA), simple network management protocol (SNMP), command line interface (CLI), simple mail transfer protocol (SMTP), XML, SOAP, Java message service (JMS) and .net. Again, a key point here is that in order to manage customers’ service levels and requirements fully, all information related to customers, their preferences and their context needs to be in the same place, normalized and accessible by the support systems as well as the service delivery infrastructure. What can you tell us about the other technology providers who are partnering with you in the cable space? How are you complementing each other? Typical technology partners are the core platform—Sun, BEA, CA (eTrust) and Oracle, who provide the infrastructure for the large scale implementations. At the edge, examples include CPE vendors (Motorola, Arris, Scientific-Atlanta, Pace, Toshiba, Texas Instruments, etc,) where we need to be able to provision to the end point for service quality and availability guarantees. In the network and services core, we will partner with or provide interfaces to all key infrastructure components from vendors including Cisco Systems, Motorola, Cedar Point, Siemens, Liberate, OpenTV, IPUnity and Openwave for example—again to meet the requirements for end-to-end fulfillment and right first time provisioning—hiding the complexity of product provisioning behind our smart portal. For whole system integration, we of course need to also integrate to the BSS components—the major billing and CRM solutions are integrated using prebuilt interfaces, or using an EAI solution via Vitria or similar. Other value-add partnerships include companies such as Camiant and Auspice. The integration with Camiant allows operators to deliver end-to-end policy-based services where QoS is needed for "pay as you go" access, bandwidth (throughput) "turbo" and IMS-based services. Auspice gives operators the ability to manage service assurance, guaranteeing customer service levels. Where the JR APS solution complements these solutions is in its ability to provide the information repository, federated, normalized and real-time. With this level of integration, cable operators can bring new services to market faster, with much better right first time provisioning rates, end-to-end QoS implementation and easier multi-vendor support. Do you anticipate an acceleration of usage-based or pre-paid products in cable’s IP arena? How would you help enable those kinds of services? Yes, we have been delivering these for a while now—for example, pay-as-you-go access, IP services and bandwidth turbo button. We provide the single point of control in the infrastructure to deliver these services and provide the self-care capabilities for users to self-select. Is there a trend away from self-installs toward tech-assisted installs? Or will these two approaches simply continue to coexist to serve different segments of the subscriber base? We see them continuing to coexist. As we move toward more (pure) IP-based services, consumers will trend to more and more self-install/self-care facilities using Web- or TV-based portals that are provided today; operators are making more of these facilities available for obvious cost savings. There are some advanced IP services that will mandate a tech-assisted install—and again, we provide a "field technician" portal to assist here where we again remove the vendor-specific configuration complexity by providing a single view portal. This also allows the field technician to upsell the customer as all these facilities are provided within the same application with the same look and feel. What is the most urgent task for cable’s technical teams in this area of converged services provisioning? Information engineering—making sure that all data related to people, preferences, infrastructure and systems is in one place, available, federated and normalized. Operators cannot deliver converged services that are clearly differentiated unless they have the information available to allow real choice in delivery for customers and speed to market of customizable, reliable services. Jonathan Tombes is editor and Jim Barthold is a contributor to Communications Technology. Reach them at [email protected] and [email protected].