AI in energy: is it as smart as you think? Part Two

By Emily Judson, Energy Policy Group, 1 April 2019

Artificial intelligence (AI) is currently in the spotlight due to recent rapid expansion of applied AI technologies across both private and public sectors. Applied AI technologies can be powerful, augmenting human capacity to address more complex situations and challenges in shorter timeframes. Indeed AI was recently named as ‘the emerging power behind daily life’ by Microsoft’s Kate Rosenshine in her keynote speech at the 2018 Tech UK Digital Ethics Summit. These new technologies also impact all levels of the UK energy system from enhancing personalisation of domestic customer service through to facilitating predictive maintenance of national transmission infrastructure.

This three-part blog series will address the growing use of AI in the energy system. Part One provided a scene-setter for readers who are unfamiliar with AI, addressing general background and terminology. Part Two (below) analyses three socio-economic enabling factors that provide a fertile ground for the adoption of AI in energy, and provides examples of where applied AI technologies are already appearing on the market. Part Three will address policy implications of the growth of AI in the energy system, opening the conversation regarding future governance.

When establishing institutions and principles supporting the governance of emerging energy technologies, policymakers must be mindful that technologies are never neutral instruments. Rather, they are human-made tools situated in context. While advances in technology can certainly be powerful, they are not guaranteed to be positive and there is a risk of unintended negative side-effects. Moving forward, policymakers must work with a wider set of actors – including industry and civil society bodies – to support the responsible development of AI in the energy sector. This work requires active, rapid development and holds the potential for significant further research.

Part 2: Why is it growing?

The expansion of applied AI has been enabled by a series of recent technical changes including: increased computational capacity (including cloud computing), increased internet-enabled mobile device ownership, increased quantity and granularity of data, and more sophisticated algorithm development. In addition to these general changes, the energy sector in particular has been affected by: cost reductions in sensors and other smart hardware (including widespread introduction of smart meters and industrial monitoring equipment), the growth of behind-the-meter services built on increased ownership of connected Internet of Things (IoT) devices, a push towards electrification of heat and mobility, proliferation of distributed renewable generation resources due to falling costs, and pressure for rapid decarbonisation. The intermittency of distributed renewable generation, which now accounts for a growing proportion of the total generation mix, has provided particular stimulus for the development of digital techniques and technologies to manage increased uncertainty. Here, the engineering challenge is shifted from load-following to demand-following and flexibility, requiring increased data and analytic capacity to coordinate a more complex system.

Alongside technical factors, the viability of AI application in the energy system is also enabled by social and economic factors, which are perhaps less discussed. Three inter-related socio-economic enablers are outlined below.

  1. Industrial policy

The UK Government’s 2017 Industrial Strategy Paper positions the AI industry as a key sector for the UK economy, in which the government advocates for expanding the UK’s ‘world-leading’ research and industry. To bolster this commitment, the strategy is supported by a targeted AI Sector Deal, and the establishment of a number of Government-supported AI and data-related bodies. These include the Office for AI, The Alan Turing Institute (the national institute for data science and AI), and Digital Catapult.

In addition, the UK Government’s Upgrading our Energy System – Smart Systems and Flexibility Plan and the Clean Growth Strategy, both support the potential to develop and use data-driven technologies to improve carbon reduction while ensuring a reliable and affordable energy supply. While there is currently no energy-specific AI body, the recent formation of the Energy Data Taskforce provide some foundations for AI-related discussions, particularly as the availability and quality of data is a key issue for AI application.

  1. Availability of funding

AI in energy occupies an advantageous position as it spans two strategic priority areas for the UK government; energy system transformation and AI development. There is therefore a lot of potential for AI R&D to receive public support. For example the AI Sector Deal outlines a total investment package of almost £1 billion in public funding for a range of activities supporting the broader AI ecosystem, including skill-building and digital infrastructure as well as direct R&D funding[1]. In November 2018 the department for Business, Energy and Industrial Strategy (BEIS) announced £84 million funding for AI, robots and smart energy innovation, including £16 million earmarked specifically for smart energy systems innovation, backed by the Industrial Strategy Challenge Fund (ISCF). The ISCF further includes a strand of £102 million in funding, designed to “make the UK prosper from the energy revolution”.

In the private sector, Venture Capitalist (VC) funds are making increasingly large investments in AI-related start-ups, on a global scale. For example, by the end of 2017, VC investment in AI start-ups had reportedly reached a total of US $12 billion, with the largest deals split evenly between the USA and China. As well as the potential for AI in energy to secure a slice of that investment pie, incumbent energy companies are increasingly investing in AI development through direct R&D funding, industry-academia partnerships, and also through tactical acquisition of AI in energy start-ups [2].  While not strictly an applied-AI business, the recent acquisition of smart storage company Sonnen by Royal Dutch Shell could be seen as demonstrating the appetite of incumbent energy companies for portfolio diversification into new digital, data-driven ventures.

  1. Social changes

The amount of time spent online has increased drastically in recent years; with UK adults currently averaging of 24 hours a week online [3]. Accordingly, and in parallel with increased automation across a range of services, tasks that used to be carried out in person are increasingly accomplished online. For example, HSBC bank now permits remote cheque deposits, where users take a photo of the document and upload it to an app. Those who use these new technologies and services may develop new behaviours and expectations as customers. One such area of change is around data collection, as typically users must provide firms with access to personal data in exchange for increasingly convenient access to a wider range of services. Users may also develop different expectations of where they expect to pay for digital services, due to ubiquity of technology business models (including Google and Facebook) where services are funded via advertising revenue, then provided for free or low cost at the point of use.

Digitalisation in the energy system is affected by the changes above. For example, connected devices such as remotely-controlled heat and lighting systems – managed via an app or virtual home assistant – are increasingly popular. Bloomberg New Energy Finance (BNEF) estimates that demand for new smart thermostats will reach 22 million per year across the USA, Canada and Europe by 2021, up from 12 million in 2018 [4].

Changes in consumer expectations are also putting pressure on energy companies to improve and diversify the ways they interact with customers, for example through the development of chat-bots, apps, user personalisation and web portals enabling users to pay bills or change tariffs online.

Another area with significant potential for growth is the automation of energy–related tasks and services within the home. For example, companies such as Labrador and Look After My Bills provide services that automate tariff and provider switching to minimise utility costs. There is also evidence that broader avenues for home automation will be important in facilitating effective use of time of use tariffs and other incentives to flatten energy demand [5]. For example, automation may prevent the need for ‘inconvenient’ behavioural change; reducing the cognitive burden associated with changing daily habits to switch energy-intensive activities, where possible, away from periods of peak demand. The potential of home automation to facilitate more rapid and reliable unlocking of flexibility at the granular domestic level marks it as an important area for future research.

Where are AI-based technologies arriving on the market?

Energy is currently a field in which applied AI products and services tend to have fairly narrow functionality, suggesting that at present applications of AI in energy are generally in the form of ‘weak AI’, running specific tasks (see Part One of this blog series for terminology definitions). AI is in general becoming relevant to an increasingly large range of tasks, spanning the full breadth of the energy system. Applications include:

  • Data analytics
    Example: Aggregation, demand management, grid planning.
    Purpose: AI can be used to detect patterns and combine insights from large or multiple datasets, including processing data in close to real time. AI may also be employed on large datasets to detect new or evolving patterns and trends, detect higher degrees of nuance, and to automate or speed up routine tasks such as system monitoring and optimisation (see below).
  • System optimisation
    Example: Distribution automation [6], grid balancing, demand response.
    Purpose: AI is used to collect, process, and draw actionable insights from larger quantities of data than can be handled economically by humans. Monitoring, analysis and optimisation can both feed into automated decision-making systems and inform human decision-making processes. For example, AI could be used to automatically optimise combined usage of wind power and energy storage, thus preventing unnecessary curtailment. However, in the event of extreme weather events an additional layer of human decision-making power may be necessary to ensure additional considerations – such as safety – are met.
  • Trading
    Example: Smart agents facilitating peer to peer trading.
    Purpose: A common use of AI in this area is to automate trades according to pre-set user preferences, pricing cues, and/or insights drawn from contextual datasets. This can happen on different scales; from household peer-to-peer trading through to algorithmic (high-frequency) trading in larger markets.
  • Customer engagement
    Example: Chat bots, voice assistants, service personalisation.

Purpose: AI has many uses in customer engagement, including learning customer behaviour and preferences to increase service personalisation. It can also drive technologies which diversify methods of human-computer interaction, for example smart speakers using voice recognition.


This blog has outlined three social and economic factors that – in parallel to technological changes – provide a supportive atmosphere for the growth of applied AI in the energy system. The blog has also provided some insight into four application areas in which products and services advertised as AI-based are increasingly seen on the market. Given the supportive socio-technical environment as above, it is reasonable to assume that the number of AI applications in the energy system is likely to grow in the immediate future. The third blog in this series will address policy implications of the growth of AI in energy, drawing from the fields of data and AI ethics to open a discussion about future governance of AI technologies in the energy sector.

[1] UK Government AI Sector Deal, 2018, page 9. Accessible via:

[2] The beginnings of this trend and conditions suggesting the potential for tactical acquisitions are identified by the financial services company BDO in their October 2017 report ‘Why Big Data, AI and Renewables are the Perfect M&A Storm’.

[3] Hymas, C. “A decade of smartphones: We now spend an entire day every week online”, The Telegraph, 1 August 2018. Available via: Accessed 8 January 2019.

[4] BNEF in Stubbe, R. “Consumers Getting Hot for Smart Thermostats”, Bloomberg Business Week, 2 January 2018. Accessible via: Accessed 19 March 2019.

[5] Paetz, A. G., Dütschke, E., and Fichtner, W. “Smart Homes as a Means to Sustainable Energy Consumption: A Study of Consumer Perceptions”, Journal of Consumer Policy, 2012, vol. 35(1), pp23-41.

[6] For further information on this process see: Tait Radio Academy, “What is Distribution Automation”, Introduction to Industrial Control Systems. Date published not provided. Accessible via: Accessed 14 January 2019.

Emily Judson is an EPSRC funded PhD student in the Energy Policy Group examining digitalisation and democratisation of energy in the context of system decarbonisation.


Supervision: Dr Iain Soutar and Prof Catherine Mitchell

You may also like...

Leave a Reply