AI for Energy Behaviour
WHAT IT IS
A team concept developed for a 2026 hackathon built around an energy challenge involving SP Group Singapore. The question we took on was not how to collect more energy data, but how to make the data a household already has legible enough to act on.
THE PROBLEM
Households can see how much electricity they use, but consumption data on its own does not explain itself. A number of kilowatt-hours does not tell you which habit produced it, whether it is unusual, or what could reasonably be changed.
THE SOLUTION
A proposed system that interprets consumption patterns and presents them as understandable insights rather than raw figures. The concept explored several levels of context (household, room, and appliance or electrical device) and how usage at each level could be translated into an interpretation of what is happening and, where appropriate, a possible behavioural recommendation.
OUTCOME
- A proposed design for turning consumption data into interpretation and possible behavioural recommendations
- Explored household, room and appliance-level context as ways of framing the same underlying usage data