Automation is Essential

Automation can reduce time-to-market, increase visibility, rescue stranded assets, and generally reduce the errors made and shortcuts taken by messy humans like you and me

The ZSM ISG’s kickoff meeting took place the second week of January, and NTT and Sprint have taken prominent roles in the continued development of the working group with DOCOMO’s Ashiq Khan and Sprint’s Serge Manning being selected of chair and co-chair of the Network Operators Council (NOC) for the effort. 

Meanwhile, the idea of Intent-Based Networking (IBN) continues to gain traction. (See: Intent-Based Networking Surges Forward in last month’s issue.) Cisco, in particular, has thrown its considerable weight behind IBN, which is really a more intelligent iteration of SDN. It contains the promise of engineers being able to focus on desired outcome with automated processes handling the configuration and execution of all of the network details.

There are countless reasons to automate, which have been discussed at length as automation has become more of a reality. It can reduce time-to-market, increase visibility, rescue stranded assets, and generally reduce the errors made and shortcuts taken by messy humans like you and me.

But it’s tough to see the central motivation as anything other than cost savings, and it would be disingenuous to imply that those cost savings wouldn’t largely be a product of reduced headcount.

I only bring this up because I’ve seen a number of pieces out there discounting the notion that network automation will render a sizable portion of the workforce obsolete. Methinks they doth protest too much.

It’s not always an easy conversation to have, but it’s not remotely unique to telecom. I’m sure most of you at least glanced at Claire Cain Miller’s New York Times piece from a year ago noting that automation is, despite protests to the contrary, a job killer. She uses the steel industry as an example, noting that between 1962 and 2005, the workforce in U.S. steel factories declined by 75 percent, the loss of some 400,000 jobs. But contrary to the popular narrative of foreign competition being behind these losses, U.S. steel shipments remained steady across that period. New technology—the minimill—meant that steel could be produced with far fewer workers.

This can be seen across every industry: self-service kiosks in restaurants and grocery stores, automated distribution centers, manufacturing robots, etc.

If you dig a little deeper, you can even see it in your own home.

According to the 1870 US census, 52% of working women were employed in ‘domestic and personal service.” Even middle-class households in the United States and Western Europe had full- or at least part-time help to handle chores such as cooking, laundry, childcare, or transportation.

But now the average modest home comes equipped with better technology for cooking and cleaning and laundry than was available to Queen Victoria herself. Our cars are easy to drive and there are no horses to shoe or barns to clean. And frankly, we’re expected to raise our own kids. We simply don’t need full-time staffs to accomplish the tasks of our daily lives.

My job isn’t safe either.

In the first year after it launched an automated article-writing program called Heliograf, the Washington Post published more than 850 automated articles. And as a big college sports fan, I’ve read my share of automated game recaps through the ESPN app. Mark Cuban famously feuded with ESPN over their use of automated game reports back in 2016, but Cuban’s fighting a losing battle on that one.

So of course there will be job losses. We can’t pretend it won’t. That’s certainly part of what’s driving the 33% cost savings that IDC reports Juniper customers achieved through automation. And Cisco claims an OPEX savings of 50-70% through its Orchestrator technology. That represents a 5-year savings of more than $65 million for a tier one operator.

But that doesn’t mean that the central promise of IBN and other automation technologies—that it frees up creative humans to tackle more complex problems—isn’t true. There will doubtless be growing pains along the way, but as 2018 continues, I expect we’ll start seeing an even clearer idea of what the future of network automation will look like.


Latest Updates

Subscribe to our YouTube Channel