AI-driven waste management for start-up Rematics

Rematics: pioneer in sustainable waste management

Rematics is an innovative start-up born from a collaboration between ACA Group and MyCSN. The young company uses advanced technologies like computer vision and AI to detect recyclable materials within waste streams when containers are emptied into garbage trucks. This provides insight into sorting quality and helps businesses separate waste more effectively, which is crucial for efficient recycling.
“Rematics aims to guide waste collectors toward more effective sorting with deeper insights into waste streams,” says Bram Peetermans, CEO of Rematics. “Our technology offers continuous monitoring and data-driven insights, helping companies make better-informed decisions and improve their waste sorting strategies."

From concept to reality: ACA’s role in establishing the start-up

Building on the results of a research and development project by Valipac, Rematics began experimenting with technologies to analyze waste in containers with the help of ACA Group. Initial pilot projects showed promising results, leading to Rematics’ official launch in 2023 with the aim of expanding its technology into a market-ready product and service package.
“In January 2024, as I joined Rematics, we thought we had a nearly final product with a suitable pricing model,” Bram recalls. “But early interactions with customers showed that both product focus and pricing needed considerable adjustments.”

In the first months, Rematics and ACA Group re-evaluated the product scope, identified key components, and refined the value proposition and development strategy based on market research.
“During this start-up phase, ACA Group proved to be the perfect sparring partner, challenging the initial product concept and brainstorming ways to align it better with market needs,” Bram explains. “The experience and expertise of ACA Group have been crucial in helping us reach our final focus.”

ACA as a development partner across domains

Market research identified that added value for potential customers lies in collecting data on sorting quality to optimize and automate processes, which requires high-quality, reliable data.
Jan Martens, Senior Project Manager AI at ACA Group, explains, “This is why we prioritized robust sensor technology that can operate reliably for 10 years in challenging conditions.” ACA quickly shifted focus from its core application development and UI/dashboarding to niche, product-related challenges, like developing custom hardware and the corresponding software.

“ACA is essentially our development partner on all fronts,” Bram says. “They assist with hardware development, manage the AWS environment, build and manage vision models, train AI, and even create our website. As development progresses, data reporting will also be used to gain better insights.”

First projects in the pipeline

The goal is to deliver the first final units by the end of 2024, deploying the hardware and software environment across multiple clients. “The current solution will offer solid basic functionality, but we also plan to learn from these projects and identify extra features that would provide added value in the market,” Bram notes.

According to Jan, “For this first MVP (Minimum Viable Product), we must continue training the model. The range of materials it must recognize has expanded significantly since the pilot phase. Eventually, we aim to develop a model capable of automatically recognizing new packaging types. We’re also working to integrate the hardware components into a single, manageable, industrial package that can be easily maintained and updated remotely from a central platform.”

The system can also issue real-time alerts to waste collectors when hazardous items, like nitrous oxide canisters, are detected, allowing for safe removal before they are compressed or sent to incinerators.