Research at Cisco

IP Based Smart Services

Project ID:


RFP-2012-074

Title:


IP Based Smart Services

Summary:


Smart Services utilize automation, intelligence-based embedded management agents and intellectual capitals to provide a proactive, predictive and preemptive service experience addressing the operations and health of the network. They turn manufacturers and others value chain intelligence of every connected device into inelegance to derive new businesses. Smart Services bends the traditional linear value chain into a "feedback loop" though which the heartbeats of manufactured objects will continually flow back though the complex business systems that create, distribute, and service those products. Adaptors of smart services are creating extraordinary performances and parries to competition underscoring the strategic impact of intelligent device networking on after slates and service management.

We are interested in funding efforts that focus on one or more of several practical problem areas for today's networks. These problems are (details are provided in the Full Description Section):

  • Systems, methods, and algorithms to capture and reuse knowledge.
  • Build support in product/network themselves including identification of new parameters (e.g. in addition the traditional SNMP, CLI output, syslog, IPSLA and Netflow data). Systems/algorithms to develop in-product capabilities to derive advanced services.
  • Fully utilize the mountain of data: Turn transactional data into usable management information that can utilized by service management solutions.
  • Define new smart services for latest technologies such as cloud, mobility, Software Defined Networks (SDN) and solutions.
  • Smarter approaches to integrate with service supply chain (possibly using open stack, open APIs, etc.)
  • Enhanced Cloud based smart service models (e.g. local data correlation vs. in the cloud), and innovative way of providing smart services using cloud
  • Enhanced Mobile based smart service models
  • Holistic management techniques and smart services for managing next generation data centers.
  • Holistic visualization techniques for observing the management data.
  • New security management and security policy techniques for new generation networks (Virtualized Data Center, Cloud, etc.)
  • Technologies to incent people to share knowledge in social networks.
  • New Algorithm to identify experts and top contributors in technical forums, social networks and crowd-sourced environments

Full Description:


Virtually all communication devices today contain a wealth of information about their status, location, usage, and performance. Until recently, this information has gone largely un-harvested and un-leveraged, even though it can offer extraordinary business benefits to the companies that manufacture, support and service those products, especially in terms of customer relationships.

As the world becomes more connected via the Internet of Things, network suppliers are asking themselves why would they minimally sell a device and forgo very essential feedback information, when they can also sell a service contract that allows them to fully monitor the actual usage and behavior of the device in the deployed environment. Usage information are not only used to service a device and prevent service deterioration by verifying contract level Service Level Agreements (SLAs) but also to determine the most essential set of future enhancements. Feedback information may be categorized by market segments but generally include common set of specific information such as features used the most, features use the least, features never used and feature usage patterns (feature A is used with feature B).

Maintaining active, open, scalable and secure channels with the device to collect feedback embedded management statistics and observes the overall device behavior in the field is perhaps the most critical value of smart services. They greatly benefit both network suppliers and customers. Network suppliers utilize the collected information to drive their go-to-market strategy that includes new and enhanced products, features, and services. Companies often uses the acquired data to drive the so-called Pareto Principle by targeting the top 20% set of enhancements first that are needed by 80% of the customers. In other words, 80% of effects come from 20% of the causes.

The advantage for customers (device owner or network administers) is even more significant. Customers can now outsource the management of their devices/networks taking advantage of Cloud Computing and Virtualization and demand guaranteed level of service agreement that includes automatic diagnostic, optimal performance and high availability. Such model has allowed customers to concentrate on their core business by leaving the infrastructure and IT supports to the experts.

Obtaining 24x7 embedded and usage information from all devices is a daunting task that can easily overload the network and overpower most advanced operation and support systems. Network suppliers are exploring new techniques to make services SMARTER by taking advantage of the new connected world. The new world driven by networked services is one in which every connected product turns manufacturers, and in many cases others along the value chain, into a new kind of 'smart service' business. It bends the traditional linear value chain into a "feedback loop" through which the heartbeats of manufactured objects will continually flow back through the complex business systems that create, distribute, and service those products. Unfortunately, while most "product-centric" businesses are now embracing the concept of growth-creating services, many are not yet embracing the strategic value driven by the shift to information-driven services. These businesses are thinking services, but they're not thinking "smart services." Early adopters of proactive connected services are creating unprecedented performance and unique barriers to competition underscoring the strategic impact of intelligent device networking on after sales and services management.

IP-based Smart Service requires identifying every managed entity with an IP address (network discovery), data collection and event correlation, based on vendor best practices and intellectual capital utilizing the Internet. Such information is used to proactively predict the network and service performance that provides information about future trends and threats to enable proactive remediation, so that network planner/administrators can take action before a problem occurs-and preemptive capabilities - those that take action based on intelligence about potential adverse events, to prevent risk-inducing conditions from occurring at all.

The most essential input for IP based service is a well defined standardized embedded measurements to be collected from the network. This includes SNMP MIBs data as defined by IETF standards for essential FACPS (Fault, Accounting, Configuration, Performance, and Security) management. When SNMP data is not sufficient, other measurements (e.g. "syslog", Netflow, IPSLA, output of Command Line Interface (CLI)) are also utilized.

The collected statistics are then consumed by various algorithms, utilizing the Intellectual Capital (IC) information(1) to calculate FCAPS and contract renewal related measures as outlined in steps 3-6 above. IC is another critical input for IP based smart services. More detailed examples are provided in Section 5.

A service becomes smart by adding software analytics to the collected and analyzed data, and delivering this results in a meaningful and actionable way that provide critical value for the customers. Smart Services provide a proactive, predictive and preemptive service experience that is automated and intelligence-based to address the operations, health, performance, and security of the network. It securely automates the collection of device, network and operations information from the network. The collected information is analyzed and correlated with the vendor's vast repository of proprietary intellectual capital turning it into actionable intelligence to aid network planners/administrators increase IT value, simplify IT infrastructure, reduce cost, and streamline processes.

Smart services enable network vendors and technology service providers to provide solutions through machine-to-machine interactions that automatically provide real-time visibility and issue resolution. Such intelligence enables people-to-people interactions and enhanced social media collaboration. The interactions enable vendors and service providers to continue growing the critical intellectual capital.

Another essential requirements for IP-based smart services is the smart agent with automated two-ways always-on connectivity between the device (or the network) and service management backend systems that typically resides in the network operation center (NOC), at the network supplier, or managing partner. This connection is used to (a) send un-interrupted near-real-time device/network intelligence from the device/network to the service management system(s) and to (b) allow network management system(s) to connect to the device/network to take action to prevent service outage or service deterioration.

Thus, one of the key differences between traditional network management and smart service is the fact that Smart Services utilizes uninterrupted, persistent machine-to-machine or machine-to-person diagnostics, fortified with intellectual capital and best practices, in a blend designed to give network administrators deep visibility into the network. Another key difference between smart services and network management being connected to a OEM (Original Equipment Manufacturer) / partner service delivery capability. Network management solutions themselves may be connected to backend services.

With smart services, network administrators have direct view and intelligence at the device, network, operations and application layer providing automated reports and recommendations. This end-to-end approach results in network intelligence that enables network vendors (typically responsible for network and service warranty), customers/clients (network owners) and partners (typically responsible for operating, monitoring and maintaining the network by working with vendors and customers) to deliver proactive services including regular monitoring, proactive notification and remote remediation to enhance the customers' network availability and performance.

We are interested in funding efforts that focus on one or more of following practical problem areas for today's networks:

  1. Systems, methods, and algorithms to capture and reuse knowledge (perhaps under the Big Data initiative)
    • Cisco receives over a million Service Requests from customers and partners per year. Most of this data is unstructured (includes text, CLI output, etc). Cisco also collects customer feedbacks, SME social network data and keeps track of quality and bug data.
    • Over 70% of customers and partners find answers on Cisco's Web by interacting with other users over social networks.
    • Advanced algorithms and systems are needed to better capture and reuse such knowledge in a very efficient way.
    • Novel approaches to incent SME (e.g. network administrators, engineers) to share knowledge and rank/ monitor contributions.
    • Objectives: accelerate MTTR, use the captured knowledge to enhance functionalities, utilize capacity and to introduce new services
  2. Build support in product/network themselves (e.g. Smart call home, Select Optimize). This includes Identification of new parameters (e.g. in addition the traditional data such as SNMP and CLI outputs as outlined above). Solution should cover the entire network.
    1. Systems/algorithms to speed up and enhance functionalities across Cisco products
    2. Systems/algorithms to develop in-product capabilities to derive advanced services (Cisco competitive advantage)
  3. Fully utilize the mountain of data: Turn transactional data into usable management information and service management data
    1. How to capture customer's data when networks are managed by partners?
    2. How to capture support data and structure the data captured from social networks
    3. How to capture and structure TAC data and share with AS and BUs to introduce new features and services
    4. How to utilize the quality data and the smart services collected raw and correlated data in this process.
    5. How to mine the captured data to validate SLA to customers.
  4. Offer new services for latest technologies and solutions
    1. Reactive: Today, Cisco Services supports latest technologies (e.g. TelePresence, security, VoIP) after the introduction of technologies
    2. Proactive (Service Automation): Develop systems/procedures to automatically support latest technologies once introduced
    3. Service Augmentation (Spiral model): Develop systems/procedures and NEW SERVICES across CA systems and practices to support latest technology before services introduction
    4. Support pre-sales service efforts
  5. Integrate with service supply chain
  6. Enhanced Cloud based service models (e.g. local data correlation vs. in the cloud).
  7. Enhanced Mobile based service models
  8. Holistic management techniques and smart services for managing next generation data centers.
  9. Holistic visualization techniques for observing the management data.

----

(1) IC information is typically captured by analyzing collected data over-time against the supplier intelligence and data bases (e.g. Microsoft collects and analyzes data from its Windows customers over the Internet).

Internet applications, collaboration tools; video streaming/conferencing and HPC applications are proliferating at an unprecedented rate. This is creating an interesting demand for the speed and bandwidth. Moving forward, this will have tremendous consequences on the design of products. While the channel density, the speed/bandwidth and number of interconnects will increase; the real estate that designers deal will remain the same in cases be reduced, in addition to more demand to lower the power consumption and cost involved in using exotic materials to improve the signal and power integrity. These contradicting needs are not easy to solve without compromises in the actual design. Moving from electrical to optical is more likely the easy path to answer all the questions and concerns. Optical interconnects using fiber are already in use and provide several advantages. The next logical trend is to do more integration using optical waveguides into PCB leading to hybrid architectures and also move to silicon photonics to enable a much more powerful chip to chip interfaces exceeding the 25Gb/s per channel. While many photonics components are available on the market to align with the industry trends; they tend to be more isolated and don't address Cisco needs.

Constraints and other information:


Cisco expects customary scholarly dissemination of results, and hopes that promising results would be made available to the community without limiting licenses, royalties, or other encumbrances.

Proposal submission:


Submit a proposal for this RFP this link will generate a new window

RFPs may be withdrawn as research proposals are funded, or interest in the specific topic is satisfied.
Submissions are batched and reviewed at the beginning of each calendar quarter in January, April, July, and October.

Questions? Contact: research@cisco.com