FAQ
Frequently asked questions
Waste Monitoring
How far can Akanthas's AI recognize the objects and materials in waste?
Our AI’s recognition capability depends on the object type, its frequency in your flows, and the quality of the captured images. Our models are already trained on the most common objects.
For more specific cases, we progressively enrich our base using images collected on your site and your field feedback. Our AI identifies material families and standard objects observable to the human eye. It does not perform chemical composition analysis (e.g., distinguishing PP, HDPE, etc.).
Like any learning system, performance evolves with exposure: the more it observes, the more it learns and gains precision. Some rarely seen materials (like asbestos) are still less well recognized, but they’ll strengthen as real examples are observed. It’s an evolving system that becomes more powerful as it watches your flows.
Is installing the cameras on containers complex?
Installing Akanthas sensors is simple, fast, and requires no works.
They’re autonomous, solar-powered, and 4G-connected — so they can be commissioned on any type of site, with no wiring or electrical hookup.
Our sensors aren’t video surveillance systems; they’re field-grade analysis tools. Their role is to objectify flows: measure fill rates, detect sorting errors, and identify risk situations.
Captured images are processed automatically by our algorithms — no continuous capture, no intrusion — and serve only to produce reliable, actionable data.
The installation is lightweight and adaptable to any environment, but its operational impact is strong: fewer reclassifications, optimized routes, and transparent operations.
How can simple alerts help us improve sorting?
An alert lets you catch an error at the right moment, correct it immediately, and hold your teams accountable before bad practices settle in. Across our clients, we see rapid improvement as soon as incorrect sorting is flagged and acted on.
The alert is only a first step: what matters is how you use it. We offer a tailored rollout for each site to turn these alerts into a true sorting management tool.
Can we see who mis-sorted?
Our solution doesn’t identify individuals — and that’s by design. It detects sorting errors automatically, without filming or recognizing people, to strictly comply with GDPR.
You can trace a zone or a flow, but never a person. The goal isn’t to point fingers but to correct drifts and build a culture of continuous improvement. We favor collective approaches that drive accountability without creating tension on the ground.
Waste Quality
Why choose Akanthas when visual characterization is cheaper?
It’s true that visual characterization can seem cheaper at first glance, especially when your teams are experienced. But the method remains subjective, untraceable, and irregular — which increases the risks of errors, disputes, and unjustified reclassifications.
With Akanthas, every characterization is: automated; reliable; documented. You eliminate subjectivity, save time, secure your processes, and get actionable evidence in case of dispute or inspection.
And once you factor in the indirect costs tied to errors, reclassifications, or unusable data, the automated solution often becomes much more economical in the medium term.
How much does Waste Quality cost?
The cost of an Akanthas solution depends on the volumes to be monitored, your objectives, the type of flows, and your technical constraints.
We scope your needs to give you a precise first estimate aligned with your priorities. We propose a solution at a fair price, truly tailored to your site, and capable of generating a solid return on investment.
Can your solution replace the MODECOM protocol?
Our solution doesn’t replace MODECOM: it complements it — and sometimes serves as an alternative depending on the use. MODECOM and the AFNOR standard remain references, but they are manual, one-off, and costly methods run on small samples.
We bring continuous, automated characterization based on larger volumes, with a high level of reliability. For many use cases (internal KPIs, EPR obligations, regular monitoring), our approach is already used as support — and even as a decision base — because it’s more responsive, less intrusive, and far more economical.
And when needed, our results can be cross-referenced with a MODECOM campaign to complete the analysis.
Waste QualiCheck
Can your solution instantly analyze flows in a sorting center?
Our solution automatically analyzes incoming flows the moment they arrive at the sorting center. It identifies non-compliant drop-offs, documents anomalies, and guarantees full traceability in case of a reclassification.
This approach lets you secure the quality of incoming loads and have objective visual evidence on hand for discussions with suppliers.
We go further by helping you durably improve at-source sorting. With our analysis data, we support your teams, your clients, and your members — to correct recurring errors, train on better practices, and drive overall sorting performance forward.
Can Akanthas really adapt to a site as specific as ours?
At Akanthas, we take the time to understand each site. No client operates the same way, and our approach is anything but standardized.
We provide intelligent sensors that we then adapt to your constraints, your flows, and your objectives. We don’t offer a rigid model — we offer a solution that truly adjusts to your field reality.
How can we be sure the Akanthas audit is reliable and truly reflects field reality?
Our solution is designed to faithfully reflect field reality.
Our sensors measure flows automatically, without human intervention, to avoid subjectivity and guarantee an analysis based on factual data. Every container or sorting zone is monitored by calibrated optical sensors that detect fill rates, sorting errors, and flow variations.
Data is then timestamped, archived, and correlated with visual proof — ensuring complete, verifiable traceability at any moment.
The audit isn’t a judgment but an objective starting point: it surfaces the losses, reclassifications, and anomalies that used to slip through. You then use those results to adjust practices, identify concrete improvement levers, and build a reliable, measurable view of your operations.
Is there a risk of the sensor being stolen, vandalized, or causing a GDPR issue?
We fully understand this concern and have planned several measures to address it.
Our system works by taking still images, not video, and the sensor is designed to be GDPR-compliant: it cannot recognize a human face or a license plate. We also provide a GDPR communication kit to install on the mast to clearly inform your teams and visitors. The sensor is positioned at around 4 meters in height, which makes it difficult to reach.
Before every installation, we run a risk assessment, and if needed, additional insurance can be set up to protect your investment. These measures combine physical security and regulatory compliance — while limiting the risks of theft or damage.