Municipal waste logistics forecasting, Jakarta Smart City

Jakarta’s waste agency serves more than 10 million residents across hundreds of collection points. Fixed schedules sent too few trucks to overflowing areas and too many to empty routes. Therefore, this mismatch increased complaints, fuel use, and emissions.

Waste at each collection point follows noisy, irregular patterns. Lebaran, Christmas, market days, heavy rain, and floods create sudden peaks. Therefore, route planning needs forecasts for individual districts and sites. The system needed hundreds of these forecasts from data of uneven quality.

I developed a forecasting pipeline that combined models, with Facebook Prophet as the primary model. Prophet made seasonal patterns easier to explain and handled holidays. Then I benchmarked it against ARIMA, SARIMA, and exponential smoothing, which model patterns in past observations. I also explored LSTM neural networks to capture complex changes over time in high-density districts. Moreover, I included public holiday calendars, rainfall, temperature, and local event schedules.

I used more than 100,000 JAKI and Qlue reports to study Jakarta’s 2020 plastic bag ban. First, I measured complaint volumes before and after the ban. Then I used difference-in-differences to compare changes over time between affected and comparison groups. This controlled for seasonal and wider economic effects, with explicit controls for COVID-19 mobility restrictions. The results provided statistically significant evidence that complaints fell after the ban.

I built Tableau dashboards showing how district waste volumes changed across locations and time. Then I translated these patterns into route recommendations for logistics planners. Moreover, I presented the results and policy findings to 500+ attendees at ICISS 2021. This was an IEEE-sponsored international conference.