Skip to main content

The path towards autonomous network operation: can you let go?

Jigsaw puzzle

Since Ford introduced the moving assembly line in 1913, organizations in many industries have had the inkling that automation might be the key not only to making incremental productivity gains but also to achieving business-model changing improvements. 

But a Ford or a Toyota only comes along once in a while and, in the meantime, automation remains like bitcoin investing: a lot more people are talking about it than are actually doing it.

For communications service providers, running large-scale, complex, mission-critical systems, the barriers to implementing network automation do indeed seem formidable.  But there is real-world proof that automation can be adopted successfully, even by similar sectors.  From manufacturing to logistics to finance, autonomous systems have revolutionized processes, bringing increases in efficiency, reliability, and scalability.  Surely it can’t be more difficult to automate even complex IP networks than it is to introduce algorithmic share trading?

This is the basis of research from the TM Forum – an industry association for telecoms service providers and their partners – which has developed a six-step maturity model for autonomous networks.  TM Forum bemoans the fact that, despite the evident benefits of automation, as of 2022, over 80% of CSPs languished at the lower levels of the model. 

Graphic showing path from manual to autonomous networks

The networking industry has achieved the remarkable feat of constructing the largest conceivable autonomous distributed networking system: the internet.  Despite that, there has been a noticeable lack of focus on involving machines in operational tasks. 

Part of the blame, we have to admit, lies with the vendor community, where there has been insufficient emphasis on operational automation, with the result that there are few solutions conducive to scalable operations across vendors. Products are often cumbersome to operate and lack a fully unified machine interface, hindering seamless integration and management of diverse network elements.

But that is not the only challenge operators face.  A shortage of skilled personnel complicates efforts to implement effective network automation.  And the absence of standardized structures further impedes progress, leaving operators struggling to navigate the complexities of modern networking environments.

The surprise, however, is that the barriers are, increasingly, not technological.  They’re human.  Partly as a result of the low-level of tool development in the industry, a culture has developed in the CSP segment that mistrusts the delegation of any but the most basic tasks to machines.  The very pride in mission-critical systems that has driven the creation of modern telecoms networks is now holding back operators from taking the next steps towards automation.

How can Nokia help?

Based on our experience and discussions with peers, it's evident that addressing several domains is essential to moving beyond identifying problems and achieving practical solutions in all areas.

Firstly, fostering a culture of automation requires developing new skills within teams.  Tools that promote collaboration and offer no-code/low-code automation capabilities can lower barriers to entry and empower individuals and teams to embark on their automation journey.  Additionally, AI tools that augment human capabilities play a vital role in democratizing access to automation.

Another critical aspect is building trust in large-scale automated systems to overcome cultural resistance.  Organizations focusing on trust-building initiatives often experience faster and more comprehensive acceptance.  For instance, digital twin technology enables teams to visualize the outcomes of proposed interventions before implementation in live systems.  Initiatives like Containerlab and Clabernetes aid in creating test environments, while using suitable Network Operating Systems (NOS) like Nokia SROS or Nokia SRLinux – designed for operation by machine – further encourages trust in automation systems.

Lastly, adopting a modular building block approach to automation, supported by structured APIs and microservices architectures, is crucial for organizational transformation.  Embracing open-source initiatives and community-driven projects not only fosters innovation but also mitigates vendor lock-in.  Leveraging technologies such as Kubernetes with schema-driven configuration facilitates a structured approach to automating systems at every level.  Ultimately, machines can assist in generating the necessary information to operate this modular system effectively.

By addressing these domains comprehensively, organizations can overcome barriers to automation adoption and reap the full benefits of automated systems.

Ultimately, however, the key to accessing the benefits of automation is not one of technology or tools.  It’s one of approach and attitude.  There are many ways gradually to build up the amount of automation used in an organization while teams become more accustomed to the new ways of working and more confident in the machines.  But the final question is: ‘Can you let go and be assisted by a machine?’

Wim Henderickx

About Wim Henderickx

Wim Henderickx is Head of Technology and Architecture at Nokia, based in Belgium.  Mr. Henderickx provides consultancy and architecture advise in various domains such as, IP networking for Fixed/Wireless and Enterprise, Cloud Computing, Automation, etc. He has over 30 years’ experience in the communications and networking industry and is a regular speaker at technical conferences all over the world. He is active in several SDO’s like IETF, etc and is an active member in open-source projects.

Wim holds a Bachelor’s degree in Industrial Engineering, Data Communications and a Masters degree in Economy and is a Bell Labs Fellow.

Article tags