The Road to Automation: How Do We Get There - Part 1

    Posted by Matt Mariani on Mar 22, 2017 2:22:41 PM

    Service providers can no longer settle for incremental improvements.

    As networks evolve to include software-based systems and virtual network functions, providers’ ability to deploy and support services improves greatly.

    In this fast-paced world, where responding quickly to the changing demands of users is paramount, service providers are required to revamp their approach to network operations.

    In this three-part blog series, we’ll examine the different pieces which will take service providers towards automation.

    The first step down this path of advancement is a matter of examining the blueprint of the network.

    Instead of continuously finding ways to make employees in the NOC or SOC marginally more efficient at alarm and ticket resolution, the goal for many service providers is to achieve lights-out operations.

    Now, wanting something and knowing how to get there are two very different things. Service providers require a proactive approach instead of a reactive one.

    It’s All in the Domain

    Building this functionality will be a challenge. Establishing near lights-out operations is entirely feasible today, but it will only happen if service providers shift their way of thinking. First, the problem needs to be approached by focusing on network domains, rather than the traditional approach of focusing on OSS functions. Second, automation must start within each domain, and not as a single loop or process across the entire network.

    The ability to automate operations across domains is important, as complex services span across multiple domains.  But implementing a control loop at this scale would be a massive undertaking due to the complexity involved. Instead, service providers should design domains to manage their own resources, and carry out auto-healing and auto-scaling individually. These processes take place behind “closed doors”, and only share the necessary output with other domains.

    The Factory Process

    Control loops such as this can already be found in a number of other industries. For example, the process to manufacture any complex product, such as aircrafts. This process usually sees multiple factories involved, with each location responsible for creating certain parts or subassemblies of the aircraft. In order to create a finished product, safely and profitably, that is ready to join the fleet, each factory has a specific objective which has to be met within a given timeframe. If something goes awry in one of the factories, a supply shortage for example, it is their responsibility to get their operations back on track, and deliver the component that is needed.

    Federating closed-loop processes into a hierarchy enables easier integration with other processes and makes the process of automation manageable. By having each domain carry out local events and it all come together seamlessly into an overarching story, this is the first step to having lights-out operations be realized, and it’s a step that can be taken today.

    Stay tuned for part 2.

    Download the Appledore Research Group’s white paper to learn how increased automation is the only way forward to support the demands of cloud services.


    Topics: Closed-loop Automation

    What do self-driving cars and next-gen networks have in common?

    Posted by Matt Mariani on Dec 6, 2016 12:56:37 PM

    Chances are if you’re the one driving during a long road trip, you’ve drawn the short end of the stick. This is because while you’re stuck paying attention to the road, those in the back are enjoying the ride.

    Uber-Self-Driving-Cars-Starting-In-Pittsburgh-Late-August-03.jpgHowever, that will all change as driverless cars are set to revolutionize the automotive industry. In the not too distant future, people will have the luxury of simply hopping into their vehicle, inputting their destination, and watching as it navigates the road autonomously.

    What’s under the hood of these driverless cars?

    This technology, which is already being rolled out in many places, including in Uber’s self driving car pilot program, is incredibly futuristic and has many are excited at the prospect of having more free time while being “behind the wheel”. The intention of this post isn’t to talk about how telecom networks support connectivity to driverless cars, but rather to consider the principles behind this highly automated system.

    This “auto-pilot” capability is achieved by combining several automated subsystems which have been gradually added to vehicles over the years: lane departure detection, adaptive cruise control, and assisted steering, for example. These are self-regulating processes, which modify their actions based on sensors and results. Automation is then achieved by identifying errors or differences in the system, and using this information to make any necessary adjustments. This process is known as a feedback control loop.


    This allows for individually automated systems to be nested on top of one another, establishing a single continuous process. This is critical, as it minimizes complexity, reduces risk, and most importantly, prevents you from ending up in a ditch.

    Telecoms shouldn’t try to reinvent the wheel!

    So this begs the question, if we can use this automation to safely transport humans, why can’t the same be done for telecom network operations?

    In the past, there has been a reluctance to automate in telecom due to the perceived risks and challenges related to bad data and complex, siloed operations. However, some of these barriers have been overcome thanks to recent innovations from the IT industry which have made it possible to process and analyse the vast amount of data produced by networks.

    A growing number of service providers have begun to integrate SDN and NFV into their network as it promises improved agility, elasticity, and cost savings. While these redeeming qualities are definitely creating a buzz within the industry, automation, as it pertains to service assurance (in addition to fulfillment), is critical to realizing the full potential of this technology.

    Automation-and-Assurance-Photo.png                                                                                                                            Graphic by Appledore Research Group

    As highlighted in the example of driverless cars, feedback control loops are the structure required for success. The concept is simple, a single and intuitive process is established which supports a wide variety of tasks, from creating and maintaining new services, to optimizing resource utilization.

    Traditionally, service fulfillment and assurance have been independent silos. However, in this new automated closed-loop environment, assurance and analytics will converge to establish an intelligent platform that guides orchestration. As networks become more complex, closed-loops provide the ability to correlate all of the relevant data, gain a comprehensive snapshot into what is happening in the network, and make a decision on what actions to take to improve operations.

    In the end, this transformation allows service providers the ability to fulfill the customized services offerings to customers, which are a key differentiator in today’s competitive market. Additionally, it allows operators to shift their focus from monitoring network performance to delivering quality customer experience.

    Learn how to speed service delivery and drive customer experience with Exanova’s ability to deliver closed-loop service assurance for vCPE at the Virtual Edge.

    New Call-to-action

    Topics: Software-Defined Networking, Network Functions Virtualization, Internet of Things, Closed-loop Automation