Intelligent Data Centres Issue 51 | Page 49

END USER INSIGHT END USER INSIGHT investment at a much faster pace . More importantly , the project was delivered within the 10-week timeframe that we initially thought it could be achieved in . We were able to achieve the targets set by the company and also look at reducing our energy going forward this year and into next year . It was a big win-win from both those objectives , and we were able to do that using the solution .
How do you plan to take advantage of EkkoSoft Critical ’ s optimisation capabilities going forward ?
Mohamed : EkkoSense has promised quarterly optimisation reports , meaning it will do a site survey or run the Cooling Advisor software as part of the solution on a quarterly basis , giving us more room for improvement . As a technology company , we constantly seek ways to modernise our legacy technology stack , which all sits within the data centre . So therefore , we have more opportunity as we start migrating some of the legacy into newer data centres . We will have further optimisation to reduce cooling . Without the EkkoSoft Critical software , we would have been unable to see and understand how the temperature profile and the optimisation can be reduced . Now with the software , we have the AI-driven insight and data points that allow us to make critical decisions and further reduce energy consumption .
How does EkkoSense ’ s software help equip Three with the real-time operational visibility needed to secure further data centre cooling energy savings in the future ?
Mohamed : The software provides real time insights so from where I ’ m sitting , I can dive straight into one of our data centres and look at how the temperature profile is set . As you can imagine , the switches , the servers , the network equipment , everything , has to be running in an optimised temperature range . Now , I can look at a particular rack , a particular aisle and know what the temperature profile is . The thing I like best about the software is it allows me to time lapse so I can go back four weeks or three months and see how the profile was at that point in time and compare . I think that ’ s a brilliant way of helping non-technology folks to understand what the differences are . The results speak for themselves – we ’ ve already started seeing a reduction in our overall consumption .
How did you help Three to meet corporate demand for 5 % total energy saving across its legacy sites ?
Boyle : First of all , it helped having a very supportive , collaborative open customer in Three , and also its partner – CBRE .
As an organisation , we ’ ve got a long history of helping customers take fairly significant energy savings out of their data centre sites . Three ’ s corporate goals were definitely something we thought we could help with . At the heart of all our deployments is EkkoSoft Critical – our software platform . We need the data inputs to feed the analytics and the software , so we put a blanket coverage of sensors across the racks and cooling units so that gives us the real time datasets we need to help deliver the optimisation savings .
The platform ’ s also got a unique analytics and Machine Learning algorithm that Shamim touched on briefly . It works across multiple datasets , crunches , millions and millions of data points and then throws out a series of recommendations . These are complex mechanical and electrical datasets that can be difficult to understand , so you ’ ve got to crunch it down into intuitive , easy to use visualisations to encourage the customer to want to use it .
That ’ s why a crucial part of EkkoSoft Critical is a piece of functionality called Cooling Advisor – a set of Machine Learning algorithms that constantly learn the environment so that it can react to change .
We take the datasets and put them through the analytics engine in the software to produce a series of very easy to understand data visualisations that allow the customer or the partner to deliver the savings . So , it could be very simple first principle saving – such as


the optimisation of airflow , the changing of setpoints , or optimisation of cooling units – and it highlights inefficiencies in the site that we can leverage and make the cooling changes needed to deliver savings . Another vital component is speed to delivery . So , we were able to deliver a fairly aggressive timescale – I think the actual first energy savings were delivered within six weeks of our project start .
Can you tell us more about how EkkoSense ’ s Machine Learning and AI-powered optimisation technology operates and why this is beneficial ?
Boyle : At the heart of EkkoSoft we have a Machine Learning component that ’ s a fundamental element of the platform . It provides a lot of functionality and a lot of often overlooked benefits . There ’ s risk avoidance operational visibility – so first and foremost , you need to provide a toolset that allows the customer to make sure there are no hidden issues in their data centres . With this we ’ re able to spot quite low level anomalies where the trends are not going in the right direction . So first and foremost , it ’ s always risk avoidance , keeping the lights on , stopping service outages and then it ’ s about delivering the payback and delivering on the goals of energy , sustainability and ESG targets . So , our Machine Learning and AI technology helps ensure risk avoidance , enables you to optimise sites and deliver cooling energy savings and also allows you to identify any stranded cooling capacity so that you can make the most of the of the available mechanical electrical workload that you ’ ve got . � www . intelligentdatacentres . com