AI is personalising energy for customers.
Here's how

Imagine a world where you can play the role of an engineer virtually troubleshooting simple problems with your boiler through augmented reality, or where your digital assistant can arrange a boiler service at a convenient time. Imagine the cost savings and environmental benefits of factories that remotely control their electricity use to avoid peak times.   

These scenarios are quickly becoming a reality through the power of Artificial Intelligence. Once limited to laboratories and science fiction, AI is now part of our everyday lives.

From simple Machine Learning that recommends what we buy on Amazon to the cutting-edge Deep Learning behind the emergence of driverless cars, AI is transforming everything we do.

Some of the most visible changes brought about by AI in recent years concern the way that customers engage with businesses.

For many, the first direct contact between a company and customer is through an AI-powered chatbot.

As the technology develops, chatbots are evolving from answering simple troubleshooting queries to executing entire online orders.

In providing responses far faster than phone calls or emails, chatbots successfully meet one of the two qualities that Google identifies as most important to consumers today; speed and personalisation.

According to a Google study last year, 63% of 18-54 year-olds in the US expect brands to use their purchase history to provide them with personalised experiences.

But the days when purchase history was the key influencer of a personalised service have long gone. The explosion of the Internet of Things and connected devices - from smartphones to smart meters and sensors - is driving a new age of data-rich customer relations, both with individuals and businesses.

There are over 23 billion connected IoT devices on the planet today, and this is expected to more than treble by 2025 (see below).

Personalising Energy

The rapid ascent of IoT-connected devices has led to the age of Big Data. And AI is being used to crunch that data to generate new insights and create ever-increasingly tailored customer experiences.

For energy customers this could mean providers like British Gas will use AI to analyse data from smart meters and home devices. These provide real-time information on the devices in customers’ homes that use the most energy, helping them find ways of becoming more energy efficient.

Energy companies can also use that learning from the data to provide recommendations on the energy package most suited to a customer’s patterns of use.

“All of that learning has been done in relative real time, so at the point that you're interacting with the customer, you can offer that service and they feel like you know them,” says Centrica CIO Mike Young.

“And it reinforces that level of intimacy between you and the customer and therefore reinforces that sense of confidence.”

Personalised recommendations are just the tip of the iceberg when it comes to how AI is improving the energy sector for customers.

From energy trading to boiler repairs, AI is transforming the way services are delivered, says Young.

“If you're trading in energy, you're using AI in trading,” says Young.

"If you're into renewables, you're using AI to fashion your best and most efficient performing renewable assets.  It’s literally everywhere. It’s driving us to where we position our engineers to go to service a boiler - the optimum route, the optimum positioning, what should they have in their van.”

The personalisation of services isn’t limited to domestic energy users.

Businesses can be offered increasingly sophisticated energy packages based on smart meter data, while major energy users can go one step further.

Factories, for example, are today fitted with thousands of IoT sensors measuring their processes, water and energy use, along with monitoring critical safety readings in real-time.

When connected to analytics software these devices help major energy users optimise their processes and become more energy efficient.

When AI is brought into the mix, factories can go even further still.  AI can also be used to control a factory’s energy use remotely, transforming it into a virtual power plant.

Virtual Power Plants

Virtual power plants take the personal approach to the next level, by learning the consumption habits of major energy users at a forensic level of detail.

Software such as REstore’s Flexpond connects to thousands of IoT sensors in factories and commercial buildings. It can capture, store and analyse gigabytes of factory data every minute.

Crucially, the AI within Flexpond also enables it to control the electricity use of these facilities in an intelligent way.

Because factories consume so much electricity, being able to control how and when they use it is a prized asset for grid operators trying to balance electricity distribution.

If a grid operator needs more capacity, it can make the request to companies like REstore - part of Centrica Business Solutions - that can adjust the demand of factories within seconds.

This relationship can be so granular, intimate and personal that the adjustment in demand can even be targeted at individual machines within a factory.

AI can learn about different factories’ processes and know instantly which buildings and individual machines can be turned down at any given time without harming the manufacturing process.

This near-instantaneous response is proving a popular alternative for grid operators, in situations where there is either a shortage of generating capacity at peak times or the available power plants take too long to fire up in order to meet the spike in demand.

And it is proving popular with major energy users as it is cost efficient.

As a result, the Demand Response market is growing rapidly in North America, the UK and Europe. In North America alone, the market is estimated to more than double in value from $8 billion in 2018 to $18 billion in 2022.

More than half of that will come from industrial users. The next largest element will come from residential premises, as smart homes and smart metering technologies combine with AI to fine tune domestic consumers’ energy use in real time.

And that is just the beginning of what AI is expected to achieve in the personalisation of the customer experience of energy users in the years and decades to come.

Virtual Assistants

The next step in personal customer service will be an evolution of existing technologies and practices.

Instead of engaging with a chatbot for simple tasks, or having to spend time on a call answering questions, imagine being able to instantly speak with a machine that sounds like a human, and knows everything about you already.

Or better still, how about having your own version of the machine to dial-up customer service on your behalf?

This is the tantalising prospect offered up by systems such as Google’s Duplex, unveiled by the company’s CEO Sundar Pichai at I/O 2018 in May.

He replayed a recording of Google Duplex calling a hairdresser and speaking with the receptionist to negotiate a time for a haircut for its user. The language is accented and full of natural pauses and responses.

Centrica CIO Mike Young says he can see such technology not only helping consumers to raise queries, but also help companies in responding to them. These types of AI tools won’t replace the necessary human touch in customer service altogether but they could help in busy periods, ensuring customers can always connect with their energy suppliers.

That is, of course, if those busy periods for energy customer service exist at all in the future. Centrica AI team lead Rawand Nawroly says broken boilers during a cold weather snap could become a thing of the past thanks to the power of IoT and AI.

“The whole concept of IoT is around devices that can connect online,” says Nawroly.

“But IoT is lacking intelligence and AI will bring that to the devices. Our Boiler IQ service is already able to remotely diagnose heating faults before telling you. In 10 years’ time, all devices at home will have chips built in. This intelligence will enable them to predict and detect faults and self-heal first before it alerts the customer or books an engineer."

Customers may already be experiencing this technology when they go to work. Panoramic Power, acquired by Centrica in 2015, supplies sensors capable of predicting when a piece of equipment in a commercial or industrial building needs repairing, simply by monitoring its energy usage.

Virtual Engineers

Back at home, when faults are detected in the future, AI will combine with Augmented Reality glasses to teach customers how to repair basic boiler faults, turning them into virtual engineers.

Nawroly says the combination of AI and Augmented Reality Glasses will lead to not only virtual engineers, but virtual doctors, counsellors and bankers too.

“The augmented reality industry is experiencing unprecedented growth,” he says.

“There are huge investments from tech giants, investing billions of dollars to build the next generation of smart glasses eyewear.”

Nawroly expects the next generation of AR glasses to appear in 2025, closely followed by the advent of the virtual engineer, bringing a new level of immediacy and intimacy to energy companies’ relationship with their customers.

All of this technology – which could even include AR contact lenses – will all work together to put customers at the centre of their world and in control, says Nawroly.

"As a customer in my home I will have technology all around me,” he says.

“Technology will be so invisible that I interact with it subconsciously and in many cases the intelligent device sensors will do all of the work for me. The device will become a companionship that I work with, to optimise my life. It will carry out the routine tasks and give me the space as a human to be creative.”

Scroll back to top