Artificial intelligence (AI), or in practical terms predictive machine based decision making based on past behaviours or actions, is being introduced increasingly in the energy sector.
Like so many other technologies making waves today, AI isn’t new but with modern automation and data gathering technologies and analytics it is more readily enabled, finding application in a growing range of use cases associated to energy and asset management.
With the move towards a decentralised, dynamic energy system, its role is clearly essential to deliver the scale and level of automation that is required.
AI use cases
Given the status of AI as an emerging technology and its potential, it is no surprise that it should be attracting the attention of tech giants such as Amazon with Echo and Google with Google Home and others looking to the booming smart consumer market.
Google also has been employing AI for its own operations. Through its AI company DeepMind, up to 40% reduction is claimed possible in the energy requirements for cooling in data centres. Inputs were data on parameters such as temperature, power, pump speeds and setpoints, which were modelled to predict the future operating conditions an hour ahead in order to control the systems more efficiently.
Major energy suppliers such as GE and Siemens also are exploiting AI in various use cases. For example, Siemens is using AI to optimise the operation of gas turbines and AI forms a key component of its MindSphere cloud offering.
Demand forecasting is an obvious use case for AI.
It forms the basis of Comverge’s demand response offering, which earlier in the year was acquired by Itron and is being integrated with the OpenWay Riva edge solution.
And earlier this year, DeepMind was reported to be in discussions with UK’s National Grid about using AI to support balancing of the grid.
“We are always excited to look at how the latest advances in technology can bring improvements in our performance,” National Grid was quoted as saying.
“We are in the very early stages of looking at the potential and exploring what opportunities [Deepmind] could offer for us.”
“We think there’s no reason why you can't think of a whole national grid of a country in the same way as you can the data centres,” said DeepMind chief executive Demis Hassabis.
Startups take on AI
With the opportunities it offers for innovation, AI is an obviously fertile area for startups in the sector.
For example, the Israeli company CyActive has employed AI in the development of a novel cybersecurity solution.
Utopus Insights, backed by IBM and Vermont Electric Power Company, uses AI for hyperlocal weather forecasting to improve the dispatchability of renewables.
Another is the UK startup Grid Edge, which has been recognized by Centrica with funding to support the commercialisation of its energy management solution which “turns buildings into batteries”.
Home energy management
Another startup which has been recognised by Centrica is Verv, a spinoff from the startup Green Running, which is focused on energy data disaggregation and analytics.
In 2015 Green Running was awarded £750,000 from Centrica’s Ignite entrepreneurial investment fund to develop its energy management technology. Out of this has emerged Verv, a ‘plug and play’ home energy ‘assistant’ which clips to a meter or smart meter and disaggregates the energy usage to the appliance level. To achieve this, Verv samples the electricity flow at a claimed million times per second.
Via an app, users can then see in real time the current consumptions and costs of the appliances in use. Verv also provides warnings of appliance faults and if a device is left on or shows signs of activity – such as lights being switched on – when away from the home, effectively doubling as a security warning system.
To date Verv devices have been bought by British Gas, EDF and the insurance company Allianz but they have yet to become available to individual purchasers, which will occur during November.
However, the company, which has garnered an impressive array of awards for its innovations, is already working on the next iteration, supported by an almost 200% subscribed crowdfunding (target £500,000, ceiling £1m). This will incorporate a blockchain transactive solution to optimise and trade energy.
Using data on the performances and use patterns of individual appliances, the system can predict how much electricity will be required by the home, as it knows what appliances are likely to be in use and what electricity they will require.
Coupled to solar panels and battery storage, the local generation potential can be forecast and any additional energy purchased at the best price, while those with an excess can trade this with a neighbour.
With support of a £215,000 grant from the government’s Department for Business, Energy and Industrial Strategy’s Energy Entrepreneurs fund, the aim is to build out a complete proof of concept of the solution.
Such concepts illustrate the power and potential of AI and the transformation it could bring to our sector. Now we need to see if consumers agree.