In environments of high complexity and potential risk being able to label components automatically can help with mutliple operational tasks.

Updating and replacing legacy analogue meters with digital meters can be an expensive and complicated process. If instead IoT cameras are positioned to capture images of the read-outs and translate those to digital read-outs then the information can be captured automatically. This enables greater speed, frequency and consistency of data capture, along with reduced risk and cost from human checking - potentially enabling multiple other use cases.

Optimise energy scheduling and management of power plants based on energy pricing, weather and other real-time data

Optimise location and building specifications in construction of power generation equipment based on previous sites and other relevant data including visual imagery (e.g. from satellites or drones). Demand projection scenarios need to be fed in to the predictive model.

Smart home devices manage consumer energy expenditure by, for example, turning lighting, heating on and off depending on human presence, pre-planned variable margins (e.g. average home temperature) and weather / time of day.

Replicate (and replace) human-made decisions in control room environments to reduce cost and human error

Optimise pricing and promotional targeting to energy customers. Factors will include issues such as weather projections, spot supply pricing and forward cost estimations.

Predict energy demand trends based on data sources ranging from sensors to social media. In a competitive market the more varied and unique the data sources accessed the more refined (if not always more accurate) the model can become.

Monitoring and indentifying potential fraud and theft from utility networks

Optimise treatment of processed materials to enable reuse, saving cost, and to reduce environmental impact, also potentially reducing legal and negative PR risk.