Featured Story
After losing Nokia, crisis-hit Intel seeks network assets buyer
Nokia is substituting Arm-based chips for Intel silicon in its latest 5G products amid talk of a possible Ericsson takeover of Intel assets.
The Norwegian telecom incumbent sheds further light on how automation and AI could help it to slash headcount and reduce costs.
Every year, as the Mobile World Congress grows its attendee numbers, the most recognizable operators thronging its cavernous halls and walkways shrink into slimmer organizational outfits.
This downsizing, as it is commonly called, claimed about 74,000 jobs at ten of the world's largest telcos between 2015 and September 2017, as companies endeavored to boost their profit margins. But ask a telco if automation and artificial intelligence will decimate an even bigger number of jobs in future and one usually gets a nervy, non-committal response. (See Efficiency Drive by Major Telcos Has Claimed 74K Jobs Since 2015.)
Proving a rare exception, straight-talking Scandinavians at Telenor Group (Nasdaq: TELN) did not even have to be asked outright. Presenting annual results at the end of January, the Norwegian telco boldly proclaimed that automation and digitization will deprive 6,000 employees of their jobs by the end of 2020. That equals about a fifth of the operator's entire 29,700-employee workforce at the end of last year. If other companies are more covertly plotting similar reductions, the outlook for many telecom-sector employees is grim.
Figure 1: Shrinking Bjorn Taale Sandberg, head of research at Norway's Telenor, says automation will happen throughout the operator's business.
Exactly where the axe will fall at Telenor has not been specified, says Bjorn Taale Sandberg, who heads up the operator's research division. Addressing analysts during Telenor's recent earnings call, CEO Sigve Brekke hinted that staff in call centers as well as IT, marketing and support functions would feel a squeeze. But the networks side of the business will certainly not be shielded from any cuts.
"The load on the networks is not falling and data growth is exponential, and we are expected to service that and maintain a margin, and maybe even grow the margin," says Sandberg in a discussion with Light Reading. "The only way to do that in the short term is to go heavily after opportunities such as digitizing the core and going for cloud-based solutions instead of lots of different data centers."
Some of this work is already underway. Like many other telcos with operations in numerous geographical markets, Telenor has previously maintained separate IT and network systems for each territory. It is now building a common network and IT delivery center for its Asian business units, where automation is high on the agenda. Telenor is also busily "cloudifying" its networks. It aims to virtualize 75% of its IT this year and to have 90% of network functions running in the cloud by 2020.
If all goes to plan, the common network and IT delivery center should lead to cost reductions of between 20% and 30% by 2020, from a "baseline" figure of about 2.6 billion Norwegian krone ($330 million). Virtualization, it is hoped, will lead to an "efficiency improvement" of 40% in Asia, where relevant "baseline" costs are roughly NOK4.2 billion ($540 million).
The overarching objective is to cut total operating costs by 1%-3% annually. Telenor easily exceeded that figure last year, when it laid off 2,600 full-time employees and slashed costs by 4.7%, to NOK44.7 billion ($5.7 billion). Its margin for earnings before interest, tax, depreciation and amortization -- a closely watched industry measure -- rose from 36.7% in 2016 to an impressive 39.3% in 2017.
While artificial intelligence will not destroy a huge number of network-related jobs in the next few years, it will become increasingly important with full virtualization and as 5G services are launched, reckons Sandberg. Telenor is already testing machine learning for resource optimization and planning. This could, in future, allow Telenor to shift software licenses between basestations to save costs.
"There are schemes where you don't pay for the full 4G capacity unless you need it," says Sandberg. "With machine learning you could move the licenses between basestations in real time and operate with fewer licenses."
Besides supporting this kind of efficiency, machine learning could aid long-term network planning by indicating where hotspots are likely to develop. In addition, intelligent networks should help Telenor to manage and deal with faults. That may also have implications for staff numbers.
Next page: Multi-vendor pain
Multi-vendor pain
But investment in AI raises questions about Telenor's arrangements with vendors such as Ericsson AB (Nasdaq: ERIC), Huawei Technologies Co. Ltd. and Nokia Corp. (NYSE: NOK), all of which regard AI as a potential sales opportunity.
For starters, there is some industry concern about the use of AI in multi-vendor networks. Through virtualization, operators hope to be able to mix and match vendors more easily and avoid being tied to one particular supplier. But ensuring that AI systems can work in this environment will not be straightforward, both Ericsson and Huawei have recently acknowledged. Telenor also thinks it could pose problems. (See Huawei Commits Up to $20B for Annual R&D, Fleshes Out AI Pitch.)
"Getting the data out of different systems with different APIs [application programming interfaces] is going to be a challenge," says Sandberg. "We are so early that we haven't really noticed this yet, but we could end up feeling that pain."
Operators and vendors might clash over AI and machine learning as they try to carve out their own spheres of influence. While Telenor is collaborating with vendors, it seems determined to build in-house AI competencies. "When it comes to the solutions and models, we don't believe in buying them," says Sandberg.
Those comments would appear to rule out certain close tie-ups with Silicon Valley giants like Amazon.com Inc. (Nasdaq: AMZN), Facebook and Google, which are pioneering the use of AI in various fields. Having previously issued a strident warning that such companies could become even more dominant unless others invest in AI, Sandberg sounds relatively sanguine when asked if Google's DeepMind could eventually worm its way into the network. (See Automation's Advocates in Downsizing Denial.)
"Google and Facebook and Amazon have a lot of data on search and translation and natural language but they don't have our network data or our customers' demographic data," he says. "Unless the big guys get access to core network signaling data they won't be able to replace operators and build better solutions."
Want to know more about cloud services? Check out our dedicated cloud services content channel here on Light Reading.
From a staffing perspective, AI will probably have its biggest impact in the customer service area. Operators such as the UK's Vodafone have already been able to reduce headcount by developing "chatbots" that can sometimes communicate more effectively than humans. Telenor is in the testing phase. "People prefer to talk to chatbots when they are good enough," says Sandberg. "As the customer preference shifts from physical to online retail, that is a big chunk of the employment and the cost." (See Chatbot Takes Charge: Vodafone's Customer Services Overhaul.)
Wherever cuts happen, investors will cheer as long as the appropriate metrics improve. Often perceived as an indicator of difficulties, headcount reductions triggered by automation and AI may be a different matter entirely. Telenor heralded a 0.7% "organic" increase in sales last year, even as it dispensed with 2,600 employees. And its revenues per employee grew 8.2%, to about NOK4.2 million ($540,000). The continued rise of that figure will be a sign that automation is paying off.
— Iain Morris, News Editor, Light Reading
You May Also Like