The utilities industry is going through a period of significant structural change and as a result businesses are embracing innovative digital technologies at a pace never seen before.
Emerging tech like artificial intelligence (AI) and machine learning, the Internet of Things (IoT), Big Data and 5G are not only seen as ways to differentiate in a competitive marketplace, but also to meet challenging regulatory and socio-economic goals, such as net zero, climate resilience and water conservation.
As the electricity grid transitions from a centralised model to accommodate multiple smaller decentralised sources of renewable energy and storage, innovative smart technologies can help monitor and manage the flow of power – and potentially provide new solutions for energy trading.
Meanwhile, the roll out of smart meters and connected homes devices including smart thermostats and voice-activated assistants are opening up avenues for consumers to cut consumption and energy bills.
Roger Hey, DSO systems and projects manager at Western Power Distribution, tells Flex: “The two major changes worldwide are around the need to decarbonise the energy system, and the meteoric rise of digital technology, which is constantly falling in price. Both are highly dependent on each other – all technology uses electricity, and the changes we’re making to the grid using things like smart sensors and IoT-type equipment are vital to tackle climate change – the timing couldn’t have been better.”
Water companies in England and Wales have committed to the ambitious goal of reaching net zero emissions by 2030, which in combination with requirements set out in Ofwat’s latest price review (including £13 billion in investment to upgrade infrastructure and services, a 30 per cent cut in pollution incidents and a 16 per cent fall in leakage within five years) has placed a keen focus on efficiencies that could be achieved using emerging tech.
Acoustic sensors, smart meters and other logger technologies are helping the sector analyse operational data from various systems in near real time. AI and machine learning technologies can accurately locate leakages remotely, as well as interrogate datasets to predict and avoid failures.
Angela MacOscar, head of innovation at Northumbrian Water, says: “Where previously, innovation was seen as a nice to have, now it is a necessity, the enormity of the challenges we face are so great.”
In the realm of customer service, 2020 will see utilities move towards greater automation and 24-7 intelligent support operations assisted by technology. Rohit Gupta, vice president and head of energy and utilities in Europe at technology consulting firm Cognizant comments: “Customer services will see the large-scale use of automation technologies such as AI and machine learning, whether to increase back office efficiencies, improve the level of self-service or provide consistent customer experience across multiple channels.
“Utilities are also expected to invest more heavily in building cognitive technologies with a view to moving towards virtual agents [a more intelligent form of chatbot].”
But the gods of technology are notoriously fickle and uncertainty surrounds what solutions will eventually enter mainstream use. A recent survey of senior industry figures, carried out by Utility Week, in association with global technology consultancy Wipro, found that over a third of technology and innovation leaders do not yet have a clear vision of the role technology will play in how they meet their industry’s challenges.
Part of the problem may be the nebulous nature of discussions around innovation which can become clouded by media hype and insufficient evidence of the benefits. The regulatory regime has also been criticised for failing to sufficiently encourage innovation or take a longer-term view.
That said, 2020 will be a year of increasing experimentation as utilities seek to move beyond pilots and testbeds to reap real transformational rewards. So what are the big areas of change and how will these develop in 2020?
Will AI and data analytics prevent leaks and outages?
Fault prediction and dynamic maintenance is a key use case for AI as utilities seek to exploit vast amounts of available smart data to more accurately forecast damage or equipment failures.
Northumbrian Water is using machine learning to process supervisory control and data acquisition data, and data from telemetry installed in its water treatment works, to predict when equipment is likely to fail.
The data is sent every 30 seconds to Microsoft’s Azure cloud services platform where algorithms process it to detect when assets are showing signs of deterioration. When a certain threshold is reached, a job is created in the water company’s asset management system instructing operatives to go out and maintain it.
Nigel Watson at Northumbrian Water says: “It’s about following a much more intelligent approach of understanding how the asset is actually performing and creating maintenance jobs accordingly, rather than just scheduling maintenance every month, every quarter or every half year when it may not be needed.”
Business consulting firm CGI worked with a European utility to develop a system that uses machine learning and AI to monitor and manage the performance of renewable energy resources, particularly wind farms, and identify any fault or pre-fault conditions. Humans are kept in the loop to examine any exceptions flagged by the system and decide if action is required.
According to Richard Hampshire, head of energy and utilities at CGI, it is helping them move from a reactive maintenance regime to preventative, proactive maintenance “with potentially significant savings”.
As part of a £71 million investment to cut leakage, Yorkshire Water is trialling a smart analytics system to identify leaks in its network around Hebden Bridge and west Sheffield. Remote monitoring specialists Servelec Technologies and water consultant Artesia Consulting interrogate real-time flow and pressure data to identify disruptions to normal patterns, which are flagged and passed to the water firm to investigate.
Northumbrian Water is doing a lot of work on developing sensors to give early warning of problems on the network. The company’s head of Innovation Angela MacOscar, says: “There will be an explosion in the water sector in terms of the amount of sensor activity going on, and 2020 is the year this will happen.”
There’s certainly much going on in this field, and we can expect to see adoption accelerate in 2020. That said, AI is only in its first phase of implementation and new layers of intelligence will be added as the technology and potential use cases for it evolve. AI might be ground breaking at this stage, but it’s certainly providing a mechanism for offering greater insight into problems like leakages.
Will flexibility platforms underpin the decarbonised grid?
As the electricity grid transitions away from a centralised model to one of multiple local networks, distribution network operators (DNOs) will require greater flexibility in electricity demand and production at a local level to help them manage constraints.
The challenge is finding enough available distributed energy resources (DER), such as battery storage, renewable solar and wind power, electric vehicles and demand response, in the right locations to balance the system.
Flexibility trading platforms that allow buyers, including DNOs and National Grid, to connect and trade energy with sellers, including prosumers or larger businesses, are considered a key element to making this vision a reality and uptake is on the increase.
Solutions include peer-to-peer marketplaces like Piclo, and platforms established by energy companies to connect direct with DERs, such as Western Power Distribution’s Flexible Power, EDF Energy’s Powershift and Centrica’s Local Energy Market. Flexibility can also be purchased from demand response aggregators, including Origami Energy, FlexitriCity and Kiwi Power.
Roger Hey at Western Power Distribution says: “Using flexibility is not just a good economic solution, it can be delivered really quickly, and can potentially buy us a bit of time until we develop a more certain picture of the future network. We seek a flexibility solution prior to designing and planning any new construction activity. Our website flexiblepower.co.uk shows all the zones our network is broken down into and where we are actively seeking service providers, including individual prosumers or businesses, to provide us with either generation or demand side response at certain times of the day, month or year, etc. ”
But there is uncertainty around how the flexibility platforms market will develop. A recent report by Ofgem concluded that too much work is being duplicated by industry parties and a more joined-up approach could realise benefits much quicker. Ofgem recently announced a plan to ensure energy markets adequately award flexibility, amend regulatory requirements to support innovation and experimentation, and form a net zero advisory group.
Energy trading pilot projects are currently under way to identify the most effective solutions. These include the £40 million government-backed Project LEO (Local Energy Oxfordshire) involving Piclo, Origami, Nuvve and EDF Energy, and two other peer-to-peer trials in different areas of London led by EDF Energy and Centrica.
“We expect to see the convergence of DSO flexibility markets starting to happen in 2020 and probably be completed in 2021,” Hey concludes.
Are homes getting smarter?
The smart meter rollout continues to be problematic. Centrica’s gamble on Hive to cushion it against fierce competition in the energy retail market has been slow to pay dividends. But with the public’s love affair with technology only growing, the domestic market can only continue to rise.
Connected homes give consumers the tools to monitor and modify their energy usage and will support the future smart grid. A recent report by PricewaterhouseCoopers forecast that UK households would spend £10.8 billion on smart tech in 2019 and indicated that the market could be increasingly lucrative for energy suppliers in the future.
Igloo Energy entered the market last year with Igloo Works, which provides smart thermostats, hybrid heating and smart electric vehicle charging. Smart thermostats control indoor heating based on weather information including outside temperature, and help identify health issues, such as poor air quality and mould risk.
Centrica’s Innovations business also invested in two smart home start-ups: Greencom Networks, whose IoT platform is used to manage distributed energy resources and enables consumers to trade flexibility, and Mixergy, whose smart hot water tank that learns the habits of homeowners so it can meet their needs without wasting heat.
For many consumers, the first practical experience of IoT will be smart speakers/voice-activated assistants, and utilities are harnessing the technology to automate the customer experience. EDF Energy’s Alexa “skill”, for Amazon Alexa and Echo devices, enables customers to use voice to check their account balance, next payment date, submit a meter reading, and verify when their tariff is due to end.
Last year Octopus Energy became the first energy supplier to offer real-time energy pricing using voice. Customers on its Agile Octopus tariff can ask Alexa questions about their energy tariff, to help them curb usage and maximise the use of cheaper energy during off peak periods.
Elderly people or those with special needs may be able to live independently for longer, thanks to an IoT-based home monitoring system developed by EDF Energy and tech start-up Howz.
The system monitors home energy usage and combines the data with sensor data for factors such as movement and temperature to work out a pattern of daily behaviour. If the pattern is broken, it alerts the individual’s family via a remote interface.
Smart meters provide basic real-time consumption information and with greater pressure from Ofgem, the rate of installations is set to accelerate. However, more sophisticated capabilities are available now. The home energy assistant Verv uses machine learning to break the mains signal down into individual home appliances.
The system uses an IoT box clamped to the mains line to sample electricity consumption at up to a million times a second and uses machine learning to analyse the specific electrical signatures of appliances. This data can help customers reduce energy consumption, alert them if a device is faulty or if they have inadvertently left it on, and enable the next step in smart grid development by promoting user-friendly demand-side response and dynamic pricing.
2020 is unlikely to be the year smart homes become mainstream, but it’s only a matter of time.
Will software robots spell the end of tiresome tasks?
Robotic process automation (RPA) is a rapidly advancing field that can help utilities boost operational efficiency by automating various high volume, repetitive, or manual tasks, either in the contact centre, the back office or field operations.
The RPA market was the fastest growing enterprise software category in 2018, according to research company Gartner, and integration with AI and other intelligence is expected drive adoption even further in 2020.
Water company United Utilities has a dedicated robotics team that currently runs around 40 processes using bots, a transformation it says has “given back” over 35,000 hours to the business to date.
Genevieve Wallace Dean, head of robots at United Utilities, tells Flex: “We use RPA for various purposes in the reporting space, from gathering multiple data sources together, through to sending text messages out to customers to remind them that they have an appointment or a visit. The robots free up people to focus on value-added activities. They help us take some of the noise away.”
Bots were used by United Utilities during the extended period of dry weather last summer to increase the frequency of operational reports used to manage and prioritise the availability of network teams. The standard one-to-four reports issued daily were ramped up to hourly updates. Another system launched by the water company in November combines human intelligence and automation to assess and update geographical data across its catchment area.
Infusing intelligence into RPA by combining machine learning capabilities with process automation, can enable bots to analyse, comprehend, and draw conclusions from data. Northumbrian Water is aiming to exploit advances in speech-to-text technology, which has improved considerably over the past year, to partially automate the call centre agent experience.
“We will look at transcribing calls with the agent in real time to see if we can automate the call wrap,” says Nigel Watson at Northumbrian Water. “The system would create automatic notes for the agent and maybe detect the overall disposition of the customer. If an operational incident is being passed into the field, our people in the field will know exactly what the customer said to the agent on the phone, the urgency of the situation and get the full context of the call.”
Will Blockchain and Augmented Reality start to deliver?
The role that emerging technologies play in the industry’s future is never certain, but recent research by Utility Week indicates that organisations are currently sceptical of the benefits of blockchain and virtual or augmented reality (AR).
While over half of senior industry figures surveyed said they had live applications for the IoT or AI, the proportion dropped to less than 20 per cent for virtual or AR, and under 10 per cent for blockchain.
Richard Hampshire, at CGI, tells Flex: “In many of the utility use cases where blockchain has been put forward, techniques and processes were already in place to effectively manage them. Blockchain, or other forms of distributed ledger technology still need to demonstrate incremental benefits over what already exists. One promising area is in the smart flexible energy system we’re heading towards.”
Blockchain could become the decentralised architecture needed to manage smart grids and related services such as flexibility trading platforms and payment systems for renewable energy and electric vehicle charging.
US-based LO3 technology has developed a blockchain-based energy platform that was recently deployed in a Local Energy Market trial run by Centrica in Cornwall. Meanwhile, energy company Ovo invested in UK energy tech company Electron to develop its blockchain-based distributed energy trading platform.
Utilities haven’t given up on AR yet either, Northumbrian Water teamed up with Ericsson and O2 this year to trial an AR of 5G.
Engineers wearing augmented reality glasses in the field will be able to consult with more experienced and knowledgeable colleagues remotely who will relay to them live information and data in
Nigel Watson, group director of information services at Northumbrian Water, tells Flex: “We’ve got older workers with a lot of knowledge, some of whom might like to have a more graduated retirement and work a couple of days a week on call from home. Using an iPad, they could help a young engineer in the field who’s dealing with an asset they perhaps haven’t seen before.”
Do we have enough data to run smarter networks?
Frequent and reliable data is the currency upon which water and power networks of the future will run, but how it should be harvested and analysed to maximise the opportunity is an area of fierce debate.
IoT sensors can track things like water leakage and water pressure, energy efficiency and energy consumption in near real time; the data can be analysed to identify ways to improve efficiency or perform more proactive maintenance.
South East Water is currently running a Narrowband IoT trial on 2,000 homes that will combine data from pressure and water quality monitors in an effort to reduce leakage levels.
But installing IoT devices across entire networks, including “dumb” legacy infrastructure with no digital capability, would require an expensive and disruptive retrofit campaign. The national grid may have hundreds of millions of data points that require sensors, which is a huge undertaking.
Rather than adopt a blanket IoT approach, a more viable alternative may be to install sensors in some areas and fill in the gaps using advanced data techniques like state estimation, machine learning and AI. “That’s what a lot of our digitalisation of the grid is now focused on,” says Roger Hey at Western Power Distribution. “Understanding which areas need what level of data estimation or measurement on them and finding ways to do it in a clever way, for example using the computing power of IBM Watson.”
The company utilises “twinning” – the practice of installing measurement devices on one part of the grid then adjusting and scaling the data based on known rules to provide an accurate estimate of how another similar area will function.
Other utilities investigating the benefits of data analytics and machine learning include Severn Trent Water, which teamed up with French company Capgemini to process five billion records of flow and pressure data in a bid to reduce leakage. United Utilities uses the AI platform HARVI to process data from weather, water demand, pump performance and electricity prices to make decisions on the most cost-effective and efficient way to run the network.
Data is the big ticket, and utility companies are only just scratching the surface of the potential it has to revolutionise operations and customer services.