Binit Uses AI for Trash Management


Early attempts at creating dedicated hardware for artificial intelligence have often been criticized as ineffective. However, Finnish startup Binit is developing an AI gadget focused specifically on rubbish—quite literally. They aim to use large language models’ (LLMs) image processing capabilities to monitor household trash.

While AI for sorting waste to improve recycling at municipal or commercial levels has been explored by various entrepreneurs (such as Greyparrot, TrashBot, Glacier), Binit’s founder, Borut Grgic, believes that household trash tracking remains an untapped market.

“We’re creating the first household waste tracker,” he tells Truth Voices, comparing the upcoming AI device to a sleep tracker but for your trash habits. “It features camera vision technology backed by a neural network, utilizing LLMs for recognizing regular household waste items.”

Founded during the pandemic, Binit has secured nearly $3 million in funding from an angel investor. The startup is developing AI hardware intended to be aesthetically pleasing and functional in the kitchen, mounted on a cabinet or wall near trash bins. This battery-operated gadget includes cameras and sensors to activate when someone is nearby, allowing items to be scanned before disposal.

Grgic explains that they are leveraging commercial LLMs, primarily OpenAI’s GPT, for image recognition. Binit tracks what the household discards, offering analytics, feedback, and gamification via an app—like a weekly rubbish score—to encourage users to reduce waste.

Initially, the team tried training their own AI model for trash recognition but achieved only around 40% accuracy. They then integrated OpenAI’s image recognition, which Grgic claims has resulted in nearly 98% accuracy.

Image credit: Binit

Grgic admits he isn’t sure why it works so well. It’s unclear whether OpenAI’s training data includes many images of trash or if the system recognizes various items due to the vast amount of data it processes. “It’s incredible accuracy,” he states, speculating that the items scanned are “common objects.”

“It can even identify whether a coffee cup has a lining by recognizing the brand,” he adds. “Users just need to pass the object in front of the camera for a bit, allowing it to capture the image from multiple angles.”

Data from scanned trash is uploaded to the cloud, where Binit analyzes it to provide user feedback. Basic analytics will be free, but premium features will be offered through subscription.

Binit also aims to provide valuable data on discarded items, which could be useful for packaging entities, assuming the service scales.

Yet, a key question remains: do people need a high-tech gadget to tell them they’re throwing away too much plastic? Don’t we already know we should reduce waste?

“It’s about habits,” argues Grgic. “We know it, but we often don’t act on it. Similar to how sleep trackers make us more mindful of our sleep, our trash tracker aims to create awareness and change behavior.”

Binit’s tests in the US showed a 40% reduction in mixed-bin waste as users engaged with the product’s transparency features. They believe this approach can help change habits.

The app will offer both analytics and information to help users minimize waste. For personalized recommendations, they plan to use LLMs based on the user’s location.

“For instance, if a user scans a piece of packaging, the app generates a card indicating the item and suggests local alternatives to reduce plastic use,” explains Grgic.

He also envisions partnerships with food waste reduction influencers.

Grgic describes the product as promoting “anti-unhinged consumption,” aligning with growing sustainability awareness. The startup advocates for replacing throwaway culture with mindful consumption, reuse, and recycling to protect the environment.

“We appear to be on the brink of a shift,” he suggests. “People are questioning if it’s necessary to throw everything away or if we can start thinking about repairing and reusing.”

Could Binit’s tracking feature be just a smartphone app? Grgic thinks it depends. While some might prefer a dedicated, hands-free trash scanner in the kitchen, the company plans to offer the scanning feature for free through their app.

Currently, Binit is piloting its AI trash scanner in five US cities (NYC; Austin, Texas; San Francisco; Oakland; and Miami) and four European cities (Paris, Helsinki, Lisbon, and Ljubljana, where Grgic is originally from).

They’re working towards a commercial launch this fall—likely in the US. The AI hardware is expected to be priced around $199, which Grgic describes as the “sweet spot” for smart home devices.

Natasha Lomas
Natasha Lomas
Senior Tech Reporter based in Europe. Previously, she reviewed smartphones for CNET UK and, prior to that, she covered business technology for (now folded into TechRepublic), where she focused on mobile and wireless, telecoms & networking, and IT skills issues. Natasha holds a First Class degree in English from Cambridge University, and an MA in journalism from Goldsmiths College, University of London.

Latest stories


Related Articles

Leave a reply

Please enter your comment!
Please enter your name here
Captcha verification failed!
CAPTCHA user score failed. Please contact us!
Continue on app