Utility Week

24 | 6TH - 12TH APRIL 2018 | UTILITY WEEK Operations & Assets Roundtable Cognizant Roundtable: Automation Tech and AI for Utilities 15 March, Charlotte Street Hotel, Soho, London The rise of the machines AI and automation has moved from being science-fiction to being an inevitability that all businesses must plan for. Alice Cooke reports from an industry roundtable exploring the issues. T he relentless march of technological progress means exciting innovations continue to offer organisations effi- ciency and productivity gains as well as ser- vice enhancements that simply couldn't have been imagined a few years ago. With utilities under constant pressure to provide smarter, more affordable services, the ability to adopt innovations is critical but not always straightforward. Embedding new devices and systems poses challenges for cultural and process change, while pioneer- ing the application of emerging technology can also carry organisational risks. These issues apply not least to the advancement of automation technologies and artificial intelligence (AI). At an invita- tion-only roundtable event at London's Char- lotte Street Hotel, sponsored by Cognizant, innovation leaders discussed the challenges and opportunities these technology fields present for their businesses. There's no doubt the scope for apply- ing automation in utilities, for example to increase consistency and reliability in trans- actional processes is ample. Some have already made inroads, but a recent report from PwC suggests there's further to go. It claims that, across UK business sec- tors, 30 per cent of existing jobs are ripe for automation. In utilities, the potential is even greater; 31.8 per cent of jobs in electric- ity and gas supply are "at risk" of automa- tion and in water a staggering 62.6 per cent of roles could be reallocated to physical or "so" robots. Round table delegates weren't surprised by the above average automation potential identified in by this report for utilities, but quickly latched on to the inflammatory lan- guage – "at risk" – which it used to describe automation opportunities. Extended discussion of the communica- tion and delivery challenges for automation project managers showed a consensus that implementing automation solutions is "a big cultural challenge" for any firm. Understand- ably, when employees find their roles are considered low skilled and transactional – ergo ripe for automation – they can be defen- sive and even obstructive. Nevertheless, as Steve Kay, head of inno- vation at Anglian Water Services, pointed out: "Automation allows us to understand how our systems are operating – it makes people's lives easier and it prevents things from going wrong – it's vital". Speaking from a retailer perspective, Charles Grey, head of technology UK at Just Energy – the parent company for Hudson and Green Star Energy – agreed. But he also advised that selling a story of AI assistance rather than replacement can help adoption run more smoothly. "Let people tell you how they feel AI can enhance their roles, not just replace them," he said. Rolling discussion on this theme advo- cated a move towards "dev-ops" approaches to automation projects, which place system developers close to current process owners, working with them to develop solutions, rather than imposing them. Turning to consider the scope for AI tech- nologies in utilities, delegates were gener- ally more tentative. AI was seen as a more advanced technology solution, and not all attendees had begun actively considering how it might advance their organisational goals. Some, however, had strong views on the relevance of AI to core business challenges, and broadly, these views fell into three camps: how AI can enhance current pro- cesses by interpreting large datasets faster than humans; how it can transform existing processes by "crunching" disparate datasets and driving towards a given business out- come – perhaps enhanced cashflow – and finally, how it might be "let loose" on com- pany and industry data to create challenges to the status quo. For Just Energy's Grey, the latter provided a particularly exciting pros- pect. "I think the real potential of AI is to ask the questions we lack the imagination to ask and come up with ways of working we lack the imagination to consider," he said. More broadly, it was agreed that AI solu- tions have enormous potential to revolution- ise customer experience, making interactions easier and more thoughtless – which was considered a good outcome, even if it drives away from "enhanced engagement", which is commonly identified as a core utility goal. It was also agreed that AI has a huge role to play in the creation of smarter energy systems, with complex balancing require- ments and an ever-diversifying set of active stakeholders. With intelligent, integrated energy sys- tems in mind, discussion turned to consider the challenges involved in bridging the gap between today's relatively dumb and reac- tive infrastructure. Smart metering, it was acknowledged, could allow water, electricity and gas to work through a single network, and allow individuals and businesses to closely regulate their usage and waste – and to "gamify" that monitoring by comparing usage to that of others. "Smart meters have the potential to trans- form the network as we'll be able to manage it better, and together," said Anglian's Kay. Though he added: "At the moment there's a layer of control in the way of that, but that may be something that can be changed going forwards, which would be a great step for all involved." Kay's comments sparked significant dis- cussion of the data-sharing barriers currently holding back the true potential of automa- tion and AI to transform utility infrastructure and customer experience. It was agreed that greater pooling of data arising from "non- competitive" but core industry processes could deliver step changes in operational and service efficiency across the sector. But currently, regulation and company reticence are preventing this pooling from happening. There's a long road ahead for utilities as they look to get the most out of automa- tion and AI solutions. But this debate le little doubt that both technology fields will increasingly define the industry's future. As Sean Heshmat, AVP venture leader at Cogni- zant put it: "There will come a point where everyone must adopt these if they want to survive."

• Cover