The Recycling Data Problem Sits on the Collection Truck

By Olivia Hart

National inbound contamination in curbside recycling runs near 17% by weight, and programs recover only about a third of the recyclables within reach, according to the Recycling Partnership. Those two numbers define the measurement problem in municipal recycling — and the sector keeps reading them after the fact, in mixed loads and year-end reports.

The argument advancing through the industry is simple: the data needed to change those numbers already exists, on the collection truck, at the address level, every collection day.

What is wrong with current recycling measurement?

Programs assess performance at two points, and both are lagging indicators. The first is the sorting facility, where inbound contamination is weighed and a load is either rejected or sold at a discount. The second is the annual diversion report — accurate to the ton, yet rarely accompanied by any analysis of the materials themselves.

The mismatch sits at the center of recycling measurement. Contamination and low participation are local and continuous; they happen one household at a time, every collection day. The default method measures them system-wide and periodically, once material is mixed and the quarter is closed.

A program cannot educate a 17% average, and it cannot aim an intervention at a diversion rate. The result is familiar: campaigns go out broad, land soft, and get judged by a number that moved for a dozen reasons. Participation, meanwhile, becomes a question only when a grant form demands a figure nobody tracks.

The takeaway: recycling data arrives as an autopsy, not a diagnosis.

Why does the collection route matter?

The collection truck is the only part of the system that meets the household where the behavior happens — and it does so repeatedly. A route passes every address several times a month, across seasons, year after year. No audit, survey or smart-bin pilot covers a service area that completely, because this coverage is a byproduct of a route the program already runs.

With AI image analysis, cameras already mounted on many collection trucks can flag contamination as each cart is emptied and tie it to the service address, with no extra step for the driver. Three signals carry most of the value:

• Set-out rates, measuring participation block by block, showing which neighborhoods improved or fell off.
• Contamination tied to the service address, turning averages into a street-by-street heat map so outreach reaches the homes creating the problem rather than every home on the route.
• Tonnage by route and material, showing where capture is won and lost long before it reaches an annual total.

Read together, they convert recycling measurement from a number that arrives late into one that arrives in time to improve.

Where has address-level data already worked?

The Municipality of East Hants offers one documented case. After it gave residents a self-service tool for what goes where and when, its recycling hotline calls fell 84% — information reaching the household at the moment of decision instead of after it. A signal that lands while it still means something is the one that changes an outcome.

The value of street-level data is not the dashboard. It is the shortened distance between noticing a problem and acting on it. A coordinator who sees contamination cluster on a few routes can send targeted outreach to those blocks and watch the rate respond, instead of mailing the whole city and hoping.

A program that measures set-out can aim a participation push at streets that have fallen off rather than spreading the budget evenly. A department that breaks diversion down by route can show a council where the tons come from — and where the program can improve. Measurement and intervention sit close enough together, at last, to form a feedback loop.

None of this requires new trucks, new routes or a pilot. It requires treating data the collection operation already produces as a performance instrument, and putting it in front of the people who run the program.

What decides what happens next?

Recycling is a numbers business now, judged on diversion targets, contamination thresholds, capture rates, grant metrics and operational efficiency. The programs that do well are the ones that see those numbers early enough to bend them.

The richest and most current source is not the sorting facility and not the year-end report. It is the collection truck, passing every cart, every week, and recording what it finds. The curb is the real scorecard.