In the era of IoT (Internet of Things), Artificial Intelligence has already proven that optimisation models can orchestrate industrial and commercial buildings from an energy consumption point of view. In the broader context of climate change, and considering that renewable energy sources (i.e., wind and solar) are variable by nature, AI prediction models can help with the energy stability issue for the grid and balance the supply with the demand. Grid integration with AI is required to deliver a reliable electricity system and assist to get to Net Zero Carbon.

There are ongoing developments in the area of Machine Learning that attempt to tackle climate change from within the energy sector. For example, from the Demand Side Management (DSM), by processing historical energy consumption and weather datasets, AI algorithms can derive forecasts of electricity production from alternative sources and use them to optimise the demand. As a result, the power generation can be maximised leading to a better-managed electricity grid. The latter can be quite diverse given the different sources of energy generation (solar, wind, geothermal, remnant fossil fuels, stored hydro). In terms of Demand Response (DR), the consumers are empowered with their own IoT-based management tools and they can shift their electricity providers and/or usage based on the energy market prices (on the basis of response times, services offered, and different business models).

In the supply chain sector, based on metrics from the International Energy Agency (IEA), transport emissions must be decreased by 43% before 2030. Freight logistics operations in Europe are struggling with ways to reduce their carbon footprints to adhere to regulations on governing logistics while satisfying the increasing demand for sustainable products from the customers. AI can empower efficient, optimised logistics through the continuous processing of streams of data captured from IoT devices in real-time. Physical Internet algorithms have proved to perform better in terms of reducing emissions and improving the logistics’ efficiency, especially when the sample sizes are large. But this would require a shift to an open global supply web, through digital twins, with real-time data utilised in simulation and machine learning models, and intelligent decision-making algorithms.