Product Announcement: Monitoring Trash on Sidewalks

Screenshot of Numina's Trash Detection Status

Trash management can be a complex challenge in cities: Cities and businesses have different systems for managing trash pick-up, bins are overfilled, and trash overflows onto the streets. In New York City, reporters have even captured our weekly struggle in Tales of Trash Hell; in its commercial corridors alone, over 4 million bags of trash hit sidewalks each year. In developing countries, waste management can be extremely costly, demanding 20% to 50% of city budgets. To help address this issue, Numina has developed a Trash App, providing real-time data and insights to help cities and businesses keep their streets clean.

The Numina Trash App allows customers to:

  • Know when trash needs to be picked up.
  • Compare trash levels before and after scheduled pick-up times and over custom time ranges.
  • Understand how trash build-up patterns change when scheduled pick-ups are adjusted.
  • Test efficacy of different waste management strategies.

Many cities and organizations follow a trash and recycling pick-up schedule, but these  schedules are often sub-optimal. Resources are uniformly allocated across all streets within a district or the city at large, despite varying trash levels and differing trash build-up patterns seasonally and sometimes on a block-by-block basis. Consistent, real-time trash-level information is difficult to come by. As one of our partners told us, “We’re coming from a place of no data. We don’t have pick-ups scheduled for the weekend, so knowing there is consistently higher trash levels on Sunday mornings is very helpful.”

Manually monitoring trash status across streets is a challenging, resource-intensive task. Some cities have implemented solutions using “smart” trash bins that can report on the amount of trash inside a bin, but these bins do nothing with regards to trash outside of the bin, whether resulting as overflow or from active littering. In some major cities, like New York City, the majority of trash is placed on and picked up from the sidewalk, sometimes directly blocking where people walk.

We heard partner cities’ pain points in efforts to optimize their waste management resources. Through engineering and design research, we built a trash detection computer vision model and the Trash App, providing insights on thousands of trash data points a day.

“We’re coming from a place of no data. We don’t have pick-ups scheduled for the weekend, so knowing there is consistently higher trash levels on Sunday mornings is very helpful.”

– Numina Partner

Numina sensors detect trash, and other objects like pedestrians and bicycles, through edge processing. Processing “at the edge” means that we do not have to send or transmit any images to analyze them; all images are analyzed on the sensor, sending only category-level, anonymized object metadata to our dashboards and the Numina API (see more in our privacy policy). Our Numina Dashboard and the Numina API provide powerful tools to understand how multi-modal traffic navigates your city streets. Now with the Trash App, we can also look at how conditions like trash build-up affect traffic behaviors and safety.

Is your city or organization looking to improve waste management? We would love to chat more and share a demo. Please email us at hello@numina.co to learn more.