Examples of AI in the Public Sector: Sustainability

2. Eliminating mental barriers to shift to solar energy

Transitioning to solar power is a critical goal for the UK as the country aims to achieve net-zero carbon emissions by 2050. While governments often provide financial incentives, such as the UK's zero VAT on solar panel installation to encourage shifting to solar energy, psychological and mental barriers still prevent widespread adoption despite well-documented benefits. 

The process of installing solar panels often appears complicated to many, and the initial cost can be intimidating despite the potential for long-term savings. Calculating the return on investment (ROI) often exceeds many people’s capabilities and comfort level. The status quo bias—the tendency to stick with what's familiar—is a powerful force that hinders good decisions. 

LLM-based assistants that help make good decisions on switching to solar

The San Francisco-based Wattbot is a Large language model (LLM)-based AI assistant that breaks down mental barriers and makes solar energy more accessible and attainable for people, supporting them in making a challenging yet one-time decision to adopt a new technology that significantly reduces their carbon emissions for many years to come.

AI assistants that use LLMs can mimic a natural conversation with the user, just like they were discussing their needs and concerns with a human advisor. Wattbot’s system provides personalised guidance, analyzes users' electricity bills, historic consumption patterns, home location, roof orientation, and weather data to accurately predict solar energy production during the year, and calculate the return on investment (ROI) for solar panel installation. 

The software can also handle simple, rule-based calculations, minimizing errors, simplifying complex pricing models, aiding users in exploring financial options and savings, and streamlining decision-making.

3. Promoting climate action on a daily basis

Every public sector service a citizen interacts with is an opportunity to use the digital space to sway their behaviour towards more sustainable options. However, many public sector services collect high-quality, valuable data but only use it internally. Using data in a consumer-facing way can help reveal and educate the population about the problem as well as set the scene for collective solutions.

Gamified experiences to nudge small habits for big carbon savings

Ant Forest, a mobile app by Chinese conglomerate, Alibaba, uses machine learning to track and analyse users’ low carbon activities (walking instead of driving, purchasing digital tickets, etc.) and tailor nudges that help reinforce these positive behaviours. 

With these activities, users can plant virtual trees in the app, thus creating a game-like environment that motivates users to participate actively in eco-friendly behaviours. Using AI, the app can recommend activities customized for each user’s preferences and past habits, increasing the likelihood of engagement. 

This way, Ant Group has successfully created a habit-building loop: residents’ contributions to sustainability are engrained in thousands of daily decisions. This solution also makes their everyday small decisions more visible and rewards small shifts in behaviour. 

4. Tackling 2038 water consumption targets with today’s technologies

According to the United Nations (UN), we’re facing an unprecedented water crisis as global freshwater demand is predicted to exceed supply by 40% by 2030. 

One part of the problem is water overuse in many countries, which means that water is consumed at a rate that exceeds the sustainable supply. The UK government, for instance, has set a legally binding target to reduce water consumption in England per person by 20% by 2038. 

Saving water is the type of activity where sustainable behaviour could be better prompted if people first saw how much water they use per activity and understood the impacts of their behaviour when using certain quantities of water. So, how might we support households in controlling water consumption? 

Machine learning to deliver the right prompts to reduce water usage

Ecolab, a global leader in water and hygiene technologies, has been using AI to improve water conservation efforts. By using machine learning algorithms, their software analyses data related to household water usage patterns and identifies areas where improvements can be made. 

Showcasing water usage, just like smart meters did with energy, will help people understand the impacts of their behaviours. Building on top of this, the system uses nudges providing prompts at the right times (like before a shower) and using social proof like leaderboards to gamify it.

Behavior design-driven AI support in sustainability

What is our take on sustainability projects that adopt AI technology in the public sector?

• AI support to help decision-making: Adopting solar power or enhancing home insulation has a profound effect on reducing an individual's carbon footprint. Despite available financial incentives, these one-time decisions can feel complex and overwhelming. AI tools can reduce the complexity of the process and offer personalised guidance, enabling individuals to make informed, impactful choices that drive long-term environmental benefits. 

AI coaching to change small everyday habits: Small, everyday actions such as recycling, reducing water usage, or using public transportation can collectively have a significant environmental impact. However, fostering consistent behaviour change requires ongoing motivation and constant feedback. AI-driven solutions can act as personalised coaches, using real-time data, feedback, and nudges to help individuals adopt and sustain these small but impactful habits, contributing to long-term sustainability goals.