The management and collection of household waste often represents a demanding task for elderly or impaired people. In particular, the increasing generation of plastic waste at home may pose a problem for these groups, as this type of waste accumulates very rapidly and occupies a considerable amount of space. This paper proposes a collaborative infrastructure to monitor household plastic waste. It consists of simple smart bins using a weight scale and a smart application that forecasts the amount of plastic generated for each bin at different time horizons out of the data provided by the smart bins. The application generates optimal routes for the waste-pickers collaborating in the system through a route-planning algorithm. This algorithm takes into account the predicted amount of plastic of each bin and the waste-picker’s location and means of transport. This proposal has been evaluated by means of a simulated scenario in Quezon City, Philippines, where severe problems with plastic waste have been identified. A set of 176 experiments have been performed to collect data that allow representing different user behaviors when generating plastic waste. The results show that our proposal enables waste-pickers to collect more than the 80% of the household plastic-waste bins before they are completely full.