Key Findings: Boards must keep a focus on ‘gamechanger’ opportunities to change the business model and or build new revenues, as well as the quicker wins in CX, operations and cost efficiency in the application of Network AI and Automation in telecoms. In addition, industry collaboration is a vital element in developing necessary scale given current economics and costs of development.
I facilitated the CXO Keynote Panel on “Delivering the business value and impact of Network Automation & AI” at FutureNet World 2024. This is my summary of our discussions.
The panel comprised senior telecoms business leaders focused on the application of AI and Automation:
Introduction: FutureNet World, London 2024
FutureNet World, a conference specialising in Network Automation and AI, was held at the Hammersmith Novotel Hotel, on the 16-17th April in London. It gathered nearly 700 participants, including leading C-level participants from telecoms operators, their technology partners, and analysts, primarily from the Technology function. The agenda and speaker listings can be found here.
I was kindly asked by Giles Cummings, CEO FutureNet, to chair the business track and the CxO Panel on “Delivering the business value and impact of Network Automation & AI” at this event. This post summarizes my observations from the panel.
Getting past the hype in Network AI and Automation
After the breathless AI tech-hype of the last couple of years, I was delighted to get the chance to facilitate the opening CxO Keynote panel.
Overall, I wanted to get a sense of how the boards of telcos now see the topic and the opportunities, to give context to the Network Automation and AI community.
Build sound foundations: earliest wins in CX, operations and costs
We started by sharing the output of the pre-event research I had conducted, shown below, which showed that the industry is much more confident of the operational and cost reduction opportunities from network automation and AI than those relating to new business models and new revenues.
The panel agreed with this perspective, but also felt that boards must give sufficient attention to developing growth potential.
Tony Gerehan, TELUS’s COO outlined how its first focus is to address the needs of its customers and regulatory and competitive pressures. He sees enhancing service reliability and availability as a priority alongside improving operating efficiency through customer self-enablement and the automation of the network and provisioning offers efficiency gains.
The company has set up a Gen AI governance board to prioritize resources on the right opportunities, and sees AI and Automation as means to reap maximum returns from its investments in telco cloud technologies.
Rick Hamilton, CEO Infovista, agreed that reducing customer friction should be the immediate priority for the application of automation and AI, pointing out that it is no secret that many operators wish to improve their customer service reputations. He said that Infovista sees a lot of its clients making strides in this direction.
Sami Komulainen, EVP Production, Elisa, added that the board needs to take a holistic perspective on what the customer gets, and the longer-term strategic journey as well as near-term ROI. In my view, Elisa is a pioneer in Network Automation, having fully automated its Network Operating Centres (NOCs) and developed the IP into a global network and automation business, including significant acquisitions such as Polystar. Sami’s view is that operators have to automate to manage the complexity of modern networks.
Initial successes
Tony cited an example of a network planning and scheduling innovation that TELUS had made that allowed a 25% redeployment of its field force through improved efficiency.
Usman Javaid, Chief Products & Marketing Officer, Orange Business, described two examples of the application of AI in new product development:
· The Threat Intelligence Platform built by Orange Cyber Defence that takes data from 5k sources and analyses vulnerabilities using AI to assess potential cybersecurity threats for Enterprises in real-time. This platform is already securing and confidently connecting 100s of enterprises.
· Supporting a coral reef restoration programme in Malaysia, where Orange provides connectivity and AI to help recognise and automatically document marine species in a protected area.
There was discussion of the cost of building and training bespoke AI LLMs, and Usman commented that Orange’s GenAI based development used open-source AI.
To make progress, be real with people and about data
Our discussion focused on data and people challenges of getting organisations to adopt AI and Automation, rather than the technologies.
The biggest ‘elephant in the room’ is people’s fear of redundancy. AI and Automation by their nature replace some human activities. The best approach is to tackle this directly – to understand that this fear is natural, to talk about it, and to develop positive ways ahead. As someone else said at the conference “you should not be afraid of AI putting you out of a job, but another employee who can use it better than you.”
Sami described Elisa’s experiences of re-skilling and re-training network engineers. He said that Elisa had learned a lot about the ethical uses of AI and security, as well as how to develop culture and skills in the organisation. Tony also talked about how new AI-based solutions can develop in 2-5 weeks, but that organisations may take much longer to adopt them. Both agreed that the only choice is to be realistic and communicate as openly as possible.
On the data availability and inter-operations issue, Usman believed that Gen AI can be an enabler to progress in the sense that it can help to cross-pollinate data from diverse sources (albeit dependent on the quality of data that the AI itself is trained on).
To avoid stagnation, invest in “Gamechanger AI”
Usman talked about how a key focus in his role is ensuring he spends as much time as he can talking about ‘gamechanger AI’ – activities that make a step change impact to the business or growth opportunities. [NB, He described the alternative as ‘everyday AI’ although I have reframed it as ‘foundational AI’ here, because while it certainly should be ‘everyday’ I feel in retrospect it could suggest that these uses are mundane or lower value, and I don’t think that is what is meant.]
I think this desire to focus on results that go beyond improving current processes is a crucial point. When the near-term benefits of a technology like AI and Automation are so strong, it may be tempting to simply shelve the harder / more speculative possibilities that will lead to growth. The longer-term consequence of systematically de-prioritizing growth experimentation is inevitably stagnation – leading to regret and recrimination.
To scale and prosper: collaborate
Network Automation and AI is complex and can be expensive. Howard Watson, CTO BT (not on this panel), described the current economic climate of the industry as “a period of choppiness – the Roaring 20s” by – and therefore less conducive to investment.
The industry’s attitude towards collaboration is therefore changing. Players perceive that by working on the right things together, they can achieve scale and progress faster.
"As suppliers, a key question for us is how much and how fast do we invest so that CSPs ability and willingness to adopt these capabilities are in line with our innovation / investment curve?" said Rick Hamilton, CEO Infovista.
“Secondly, there are opportunities to collaborate on the access to and protection of data, standards on regulatory compliance, data accessibility, and the "interface" between OSS / BSS etc., and AI. I also think the current "context" of APIs will have to advance dramatically - how do we define that in a way that everyone is not ‘rolling their own’?”
To conclude: some realistic grounds for optimism
Network Automation and AI are undoubtedly major sources of value in telecoms. The temptation will be, as always, to apply these technologies to do more of the same ‘let’s get better at what we do now!’
This is not wrong – it is necessary but not sufficient. Truly transformational and innovative companies will also use Network Automation and AI to create new propositions and offerings to customers.
I like Usman’s Orange Cybersecurity proposition as an example of this – using AI to help analyse and evolve a complex, data-rich field to benefit customers in new ways. In this case, the SP already has both IP and data that enable the proposition. This is one possible template for success.
It is still early days, and I would like to see more examples like this. I look forward to hearing about and finding the others that exist and will evolve soon to pave the way ahead.
My thanks to such an excellent panel, an expert audience, and to Giles Cummings and the team at FutureNet World who did a wonderful job of pulling such a great crew together.
Comments