Atlanta Airport's AI Bins Boost Recycling Rate by 61%

Atlanta's waste diversion problem predates Oscar Sort by more than fifteen years. In 2009 the airport introduced GreenSort ATL, an attempt to separate recyclables from general waste after collection through single-stream sorting, but contamination made the material too degraded to recover economically. The airport later ran multiple rounds of requests for proposals for a facility code-named Green Acres, intended to sort waste on airport grounds, though no contractor ever built it. Both efforts targeted the back end of the waste stream rather than the point where passengers actually make disposal decisions, a distinction airport officials now describe as the reason neither initiative moved the needle on the city's landfill diversion target.

Sustainability analyst Malaz Alhirsh said the data now flowing from the Oscar units is proving valuable specifically because it gives airport officials real-time visibility into what passengers are discarding and how they respond to different prompts, something neither the 2009 experiment nor the unbuilt sorting facility could have provided. With 250,000 direct interactions logged since July, the airport has more granular waste-stream data in a few months than either of its earlier initiatives generated over years, information Alhirsh said is already shaping conversations about the signage and rules the airport sets for its food-court concessionaires. That balancing act between federal aviation oversight and city council waste policy is why Alhirsh said real-time data now beats periodic audits for deciding what changes are worth making.

 

Read more: Ampotech Wins Special Recognition Award for AI-Enabled Decarbonisation

 

Oscar Sort Targets a Behavioral Problem Rather Than a Sorting Problem

 

Intuitive AI's founders, who grew up in India and Pakistan, said they set out to build the technology after watching how waste exported from wealthier countries accumulated in the developing world, and began experimenting with robotics and AI approaches roughly a decade ago. Their first design was a sealed unit that removed the human sorting decision entirely, but co-founder Hassan Murad has said that approach failed to address the underlying behavioral problem, since sorting decisions are typically made in about three seconds and get the wrong bin roughly 70 percent of the time. Part of that error rate stems from fragmentation rather than carelessness, as the United States alone has more than 9,000 distinct local recycling regulations, meaning the correct answer at one airport bin can differ from the correct answer at another terminal entirely.

Oscar Sort instead keeps a person in the loop, using a screen and camera to identify each item a user is about to discard and prompting correct disposal in real time, with positive reinforcement for right answers and a more critical response, styled after its namesake Sesame Street character, when items like a multi-part coffee cup are not separated correctly. The system also runs its recognition models on the device itself rather than continuously querying the cloud, and the company says it does not capture facial images, a design choice aimed at limiting both latency and privacy concerns in high-traffic public spaces. That combination of immediate feedback and light gamification is what the company credits for a secondary effect beyond the airport's own diversion numbers, with children in some terminals reportedly picking up loose trash specifically to engage with the units rather than treating disposal as a chore.

 

Explore OneStop ESG Marketplace: Circular economy & recycling

 

A Growing Network Effect Across Airports and Global Events

 

Beyond Atlanta, Intuitive AI has deployed Oscar units at Boston Logan, Winnipeg, John F. Kennedy, LaGuardia and Phoenix airports, as well as university campuses, corporate offices and sports venues, with each new site adding to what the company describes as a three-layer learning system spanning individual units, airport-level dashboards and a global data layer. That architecture allows a decision learned from a new type of waste item encountered in one country, enterprise sales lead Corey Morrow said, to be incorporated into the localized decision logic used at an airport bin elsewhere, with regulatory changes pushed out through remote software updates rather than hardware replacement. Airport officials have also become an informal referral network for the technology, with Atlanta's sustainability director learning about Oscar Sort from counterparts at Seattle-Tacoma International Airport at an industry conference, despite what Morrow described as an aviation-sector adage that every airport operates differently enough that lessons rarely transfer directly.

The technology gained additional visibility this summer at World Cup venues across the United States and Canada, where the same touchscreen interface has been adapted to carry advertising and fan engagement content, a revenue feature that has drawn interest from beverage companies including Coca-Cola and Pepsi supplying units at venues where they hold pouring rights. That advertising layer gives Intuitive AI a second revenue stream beyond hardware and software sales, linking its growth to two different kinds of client relationships, long-term airport infrastructure contracts and short-term sponsorship-driven event placements, both feeding usage data back into the same global system.

 

What to Watch

 

Whether Atlanta's 61 percent recycling improvement figure holds up as more of its nine concourses receive units, and whether that translates into measurable progress toward the city's 90 percent landfill diversion target, will be the clearest test of whether Oscar Sort succeeds where GreenSort ATL and the unbuilt Green Acres facility did not. The advertising and sponsorship revenue model tested at World Cup venues this summer is also worth monitoring, since additional sponsor interest from companies with pouring rights could help offset deployment costs at airports weighing whether to expand beyond an initial pilot batch of units. Morrow has framed the potential impact in broader terms, saying a 10 percent improvement in recycling rates would be roughly equivalent to removing about 2 billion cars from the road, a comparison that underscores how much of Oscar Sort's environmental case rests on incremental behavioral change accumulating across a large number of small, three-second decisions rather than any single large-scale fix. How quickly Intuitive AI can convert Atlanta's early usage numbers into a case study for other major hub airports considering their own pilots will determine whether the company's growth continues to spread through the same informal, conference-driven referrals that brought Oscar Sort to Atlanta in the first place.