Emergence of shared behaviour in distributed scheduling systems for domestic appliances

When energy prices change during the day, users will schedule their appliances with the aim of minimizing their bill. If the variable price component depends on the peak demand during each given hour, users will distribute their consumption more evenly during the day, resulting in lower peak consumption. The process can be automated by means of an Energy Management System that chooses the best schedule that satisfies the user's delay tolerance threshold. In turn, delay tolerance thresholds may slowly vary over time. In fact, users may be willing to change their threshold to match the threshold of their social group, especially if there is evidence that friends with a more flexible approach have paid a lower bill. We show that social interaction can increase the flexibility of users and lower the peak power, resulting in a more smooth usage of energy throughout the day.