CSPs: The power of the energy data cloud
The need has never been greater. Communications Service Providers (CSPs) are responsible for 2-3% of global energy demand and as data-intensive applications and access to mobile broadband increases, GSMA estimates that even in a best case scenario, mobile networks will at least double their power consumption (GwH) between 2020 and 2025. While demand increases, energy price hikes hurt CSPs’ bottom line at a crucial time when revenue growth is scarce, and capital borrowing costs are increasing.
Technology is an important enabler in solving this conundrum; 5G is the first wireless technology incorporating energy efficiency from first principles - both from next generation radio antennas, and with the paradigm shift to software-defined networking.
At Google we care deeply about sustainability. We are one of the largest energy users in the Internet economy. In 2017 we became the first major company to match 100% of our annual electricity use with renewable energy purchases, and we’re committed to achieving net-zero emissions across our operations and value chain by 2030.
Our ability to optimize and decarbonize has been fuelled by our parallel mission to be an AI-first company. A big part of our third decade of climate action is not just reducing our own footprint; it’s working with our customers and partners to achieve their ambitious goals. With 9 out of 10 of the top CSPs globally trusting their data with Google Cloud, we’re excited to see how we can partner with CSPs on this journey. From our global work, we’re seeing a couple of approaches:
CSPs are rightfully focused on optimizing their Radio Access Networks; typically 73% of mobile operators' total energy consumption. CTOs are increasingly looking beyond Customer Quality of Experience, to baseline power consumption, employing a couple of methods:
A. Hardware-focused through site simplification, switching off of previous power-hungry generations, radio antennae upgrades, etc.
B. Evolving RAN as sensor; AI-driven event management to proactively augment network power consumption in real-time to meet forecast traffic load and provide ‘just enough’ cell site capacity to maintain customer experience, whilst dramatically reducing consumption.
We’re seeing multiple CSPs deploying AI/ML for this purpose, sometimes working with Network Equipment Providers to leverage their energy saving SaaS, or managed service solutions to optimize their stack. We are also seeing CSPs partner with Global System Integrators to develop vendor agnostic AI/ML that can integrate with the OSS, and feed energy saving data commands to the Element Management System (EMS). The latter, supporting CSP data scientists to incorporate custom AI models, leverage the benefits of public cloud, and drive interoperability standards across equipment vendors and more consistent energy management KPIs.
We see CSP leaders striving to build a holistic picture of energy and carbon across their estate; RAN, Core, IT, Corporate (offices/retail/fleet), often to serve ESG reporting. In Enterprise IT, CSPs can recognize a sizable sustainability dividend by moving workloads to the cloud. For example Vodafone decommissioned over 350 servers and based on Google data, are saving 700 metric tonnes of carbon a year, the equivalent to 5,900 houses, and are saving 700 metric tonnes of carbon a year, the equivalent to 5,900 houses, from their recent migration of their SAP estate to Google Cloud.
However, understanding energy consumption across the entirety of the CSPs’ business is challenging and data can often be patchy. We see major operators in Europe using AI tools to easily ingest and decode their electricity bills, verify their consumption and optimize their tariff. CSPs are also working to standardize this data across disparate OpCos - organizing, storing it, and serving it up in a reliable, timely way to business users.
Looking ahead, we see these two approaches coming together to build an actionable ‘energy data cloud’ - a standardized data repository of real-time energy consumption to support increasingly closed loop, autonomous operations. What's more, when combined with asset inventory, the energy data cloud can unlock a raft of use cases around more capital and energy efficient network planning and deployment, and optimization of field force and commercial fleet. Or by extending the process into energy procurement, leveraging data to buy more economically or when the grid is at its cleanest, to support 24/7 carbon free energy.
Beyond efficiency, at Google Cloud we're always thinking about partnering with CSPs for growth. In addition to the measurable brand and consumer loyalty uplift that operating a sustainable business brings, we’re seeing growing interest from CSPs in working with us to package sustainability within new products and services, for example Sustainability Insights as a Service to SMB customers to lighten their load on ESG reporting.
The time is now, and with a double financial and societal imperative, the opportunity is clear.