Digital Twins: A New Tool in the Network Management Arsenal

In the Austin Powers movies, Dr Evil creates a miniature clone (Mini-Me), identical in every way but one-eighth his size. In a case of life imitating art, digital twins are now becoming widespread tools in various industries "to optimize the operation and maintenance of physical assets, systems and manufacturing processes." A digital twin can be used for monitoring, diagnostics and prognostics. Typical applications include aircraft engines, wind turbines, locomotives and buildings.

According to Professor Tim Broyd, former president of the Institution of Civil Engineers, "By involving many more design, manufacturing, and asset management parties to collaborate at a much earlier stage than usual, the use of digital twins has cut the cost of delivering capital built assets by around 20% over the last five years."

Arguably, the digital twin concept has been applied to telecom networks for years in the form of monitoring software commonly found in a network operations center. However, there is perhaps scope for network emulation and sandbox modelling to be taken to a higher level by applying the digital twin concept.

According to Craig Badrick, CEO of Turn-key Technologies, "Digital twinning provides corporate IT teams with a low-risk, high-reward environment where they can experiment with novel solutions. In addition to facilitating better network design, a digital twin of a network gives an IT team the ability to run simulations for any event imaginable -- a rapid influx of network connections, a specific kind of cybersecurity breach, etc. -- and adjust its actual network configuration as needed."

Ganesan Arulanandham, consulting partner at Wipro Ltd. (NYSE: WIT), waxes lyrical about the merits of digital twinning. "DT is the melting pot of many of the latest technologies including big data analytics, artificial intelligence (AI)/machine learning (ML), immersive experiences, the cloud, sensors, open standard APIs and 5G technologies, all of which are available to take telecommunications into the much-desired digital future. DT could help providers intelligently design their services and networks and with its proactive monitoring and predictive maintenance functionalities it could potentially put an end to customer complaints."

Arulanandham sees several applications for the digital twin concept in telecom, including the following:

  • Tower management/field service management: Various data including proximity, image, touch, temperature, motion and position, can be collected from telecom sites using sensor networks. This data can be fed into a digital twin of the tower to give operations and field service management key information before they go on site. When on site, experts can also assist field workers from the command center by observing the digital twin.

  • Network planning and design: Maintaining an accurate inventory of network elements and keeping track of changing configuration has always been a challenge for operators. CSPs use a variety of tools in network modelling, planning, simulation, deployment and operation support activities. A digital twin could bring together all these tool capabilities to provide accurate network inventory and user data from live operations.

  • Programmable network DevOps: Every new technology wave (e.g. SDN/NFV, 5G) requires testing the interworking of multiple vendor devices and solutions. A digital twin of the network and associated services together with all functionalities and behaviors could become the DevOps sandbox, where new services are simulated, tested and adjusted before being deployed on the real network.

    According to Andrew Burrell, head of marketing for ultra-broadband and analytics services at Nokia Corp. (NYSE: NOK), "When it comes to network planning and optimization, a digital twin goes far beyond current software tools. Unlike traditional techniques, it incorporates everything that affects the network including data about where and how people are using it, as well as trends gathered from social media. Everything can be used to feed into these simulations -- the physical environment, usage at different locations, mapping, buildings, vegetation lifecycles, historical weather conditions and the location of street furniture."

    Although enthusiastic about the promise of digital twins in networking, Badrick notes "not only is mature digital twinning technology still several years away, but once it becomes commonplace, it will still take a fair share of networking expertise to deploy it effectively."

    There is also scope for the digital twin concept to be applied at a higher level in the organization, enabling performance of the business to be monitored and providing analytics for greater operational intelligence.

    Gartner Inc. defines a digital twin of an organization (DTO) as "a dynamic software model of any organization that relies on operational and/or other data to understand how an organization operationalizes its business model, connects with its current state, responds to changes, deploys resources and delivers expected customer value." Gartner notes that a DTO is not a replacement for business intelligence or ERP. Nor is it even a technology or a product. Instead a DTO is about organization, business operation skills, methodologies, metrics and governance. Leadership, culture and people are critical success factors.

    Unlike Mini-Me, digital twins are for good, not evil. I expect them to be a bigger part of the discussion around network and service management going forward.

    — James Crawshaw, Senior Analyst, Heavy Reading

  • ChiefMar76368 8/10/2018 | 10:21:54 AM
    Borrowed idea - but with potential Thinking around the digital twin concet in telecom is for the most part borrowed from industrial context, with the same intent: avoiding having to send a person into the field to diagnose a problem (see JCB's example). In fact the concept has much greater potential when applied to network as a whole, or indeed to services. Since network change is always a delicate business, a digital twin that permits changes to be evaluated, experimented with, tried out and validated (not only at a box level) would seem to be of real value. Even if (or perhaps even especially!) if that change becomes entirely software controlled. 
    Sign In