Frequently asked questions
Start here if you've got a question about our technology.
If you can't find the answer you need in our FAQs, reach out to a member of our team below.
Start here if you've got a question about our technology.
If you can't find the answer you need in our FAQs, reach out to a member of our team below.
Greyparrot Analyzer gathers data on waste passing over conveyor belts in materials recovery facilities (MRFs) and plastics recovery facilities (PRFs). We don’t currently support installations for collections vehicles or smart bins.
Learn more about our solutions for recovery facilities here.
Yes. Analyzer units are designed be deployed over conveyor belts. Some of our partners have integrated Analyzer data into in their own machinery using Sync.
Learn more about Sync here.
Analyzer can currently identify 111 different kinds of solid municipal waste, with categories for hard and flexible plastics, fibre, metal, waste electronics and composite packaging.
Our taxonomy is designed to work for a wide variety of use cases but we're looking to further expand into other domains.
If you have any questions about our recognition library — or want to request that we add a new material — please reach out here.
See everything that Analyzer can recognise here.
We don’t produce smart robotics. Instead, Greyparrot Sync integrations seamlessly connect real-time waste data to new and existing machinery and software, regardless of their brand.
Learn more about integrations here.
We offer a consultative approach, meaning we don’t just deliver a product or service but we act as strategic partners to find the best route and scope the needs for each customer.
Our team combines top-tier technical expertise with hands-on waste sector experience with colleagues who worked with Suez (Hannah) and Veolia (Dom) previously, allowing us to solve real-world problems with practical, scalable solutions.
Our team of experts is deeply committed to our customers’ success, and our customer service goes beyond standard support to truly integrated, long-term collaboration from initial set up to go-live and on-going support on new features.
Greyparrot is also culturally diverse which gives us a deep understanding of the different markets and local specificities.
The company’s headquarter is indeed in London, however we have team members based in Spain, France and the Netherlands. We have more than 15 nationalities among the 50+ employees in the company. This diversity in terms of culture, languages spoken, and market understanding is a real differentiator.
We work with trusted partners, such as Bollegraaf in the Netherlands, NextWaste in France, Rixera in Italy, and Van Dyk in North America, and ACI in South Korea who bring deep expertise in local markets, industry practices, and regulatory environments. Our Analyzer is live from Mexico to Australia, operating across nearly 20 countries spanning West to East — a testament to our proven scalability and international reach.
Our AI runs directly on the Analyzer at the sorting facility using which means all processing happens locally without constant data transfer to energy-intensive data centers. This dramatically reduces energy consumption compared to traditional AI systems. While we do train our models off-site, the scale is minimal compared to general-purpose AI technologies, making our environmental footprint far lower and more controlled.
Edge AI models doing image classification or sensor analysis often draw similar or even less power than gaming laptops or consoles. Example in the article: recycling facility diverted 66,000 tons of waste and avoided 20 tCO₂eq emissions using edge AI on mounted cameras.
While our AI models are optimised to run efficiently at the edge, we recognise that our hardware still consumes energy and involves the use of critical materials. To address this, Greyparrot is actively improving the sustainability of its products. For instance, the latest Analyzer units are built from lightweight, recyclable materials and consume 10–15% less energy than the previous generation, thanks to a more efficient GPU and hardware design. These changes are part of our broader commitment to reducing the environmental footprint of our technology. More information on our latest efforts can be found here:
https://www.greyparrot.ai/build-collective-backs-greyparrot-analyzer
We have several examples where Greyparrot enabled the operations manager at the facility to make significant savings.
For instance, a US MRF used Analyzer to monitor the impact of recently-purchased sorting machinery – and revealed that it was causing purity to drop from 93% to 80%. The drop cost them £800 each day or £240k over the course of a year. Analyzer data enabled the facility manager to request crucial machinery recalibration.
Another example, for one UK PRF, Greyparrot Analyzer recorded £1.6m worth of valuable material being lost to landfill each year. This drove targeted investment to recapture that value.For Green Tech, the objective was to increase throughput and maintain quality. Analyzer were deployed covering infeed lines in facilities across Romania, Slovakia & Lithuania. As a result, an increased material throughput by 200 tonnes per month using data-driven infeed blending and alerts feature to spot trends for performance improvement.
For Cheshire West Recycling, the goal was to maximise plant efficiency. Analyzer were installed on infeed, product lines at MRF to monitor operations. Actionable data were gathered in just 6 days. Using the belt status feature the Analyzer revealed a 40-minute disruption in a key product line due to a delayed haulier, prompting a review of yard staffing and the potential expansion of haulage services for better operational control.
For Re-Gen, the objective was to meet customer product purity requirements. Analyzer were deployed covering output and residues lines in MRF 200,000 tonnes pa. A rapid response ensured the final output remained on-spec by addressing contamination issues in real time, reducing response times by 24 hours.
Finally, when a Greyparrot Analyzer alert flagged contamination on an aluminium line, managers at a European MRF diagnosed the issue in under 12 minutes, instead of 38h later during manual spot sampling. This swift action prevented contamination and saved £47k by avoiding the need to reprocess 1.6 days’ worth of production.
We have multiple testimonials from customers such as GreenTech, Cheshire West Recycling and Re-Gen using our dashboard to get data and convert them into actionable insights to drive efficiencies in their facilities.
AI waste analytics uses artificial intelligence to analyse waste streams in real-time. Cameras capture images of materials on conveyor belts, and AI waste analytics systems like Greyparrot Analyzer identify the materials. That real-time data is then displayed on live dashboards, accessed via a web portal.
Learn how it works here.
Waste intelligence is the actionable insight that facility managers extract from Analyzer’s detailed waste data.
Waste intelligence insights enable facility managers to adapt as material flows and market conditions shift, operating leaner and more efficient recovery facilities.
Learn more about waste intelligence here.
The method our AI uses to predict a waste object’s weight depends on the the object’s waste class:
For item-based waste classes (e.g. aluminium cans), our AI multiplies the number of items detected by the average mass of those items. If the average can arriving at the facility weights 15 grams and Analyzer spots 10 cans, the total mass of those cans in the waste stream will be 150 grams.
For area-based waste classes (e.g. metallised film) our AI identifies the area of the material, then multiplies it by the average mass-per-area of that waste class. That average varies from facility to facility, so we establish them at the setup and calibration stage.
Learn more about how our AI waste analytics system works here.
Waste objects that don’t fit into any of our current classes are organised into one of our general material categories.
You can still generate reports on these items in the same way you would any other class.
Currently, our generalised classes are:
See everything that Analyzer can recognise here.
Analyzer isn’t limited by throughput rates. Instead, our system’s performance depends on how visible your material is.
When there are multiple layers of waste on a conveyor belt, our cameras won’t be able to see all of your material — though our calibration process enables it to make accurate estimations about composition.
Our AI can identify any object that is large enough to be identified with the human eye. Any item too small to identify will be considered “fine material”, which generally includes objects smaller than 30mm.
Yes. Our AI uses visual features to identify the function of waste objects (e.g. a plastic yoghurt bottle) and accurately determine whether it is food-grade or non-food grade.
Our customers use Greyparrot Analyzer data to actively monitor facility performance, optimise infeed blends, adjust throughput rates, track product quality, reduce the value they lose to residue lines, and more.
Every facility has its own unique application for Analyzer data. As an AI partner, we help our customers identify the best use-cases for our system with our dedicated customer success and data teams.
Learn how some of our customers use Analyzer data here.
The Analyzer portal is accessible from any device with an internet connection and a screen.
It features live dashboards that translate raw data into actionable insights, surfacing key performance indicators like sorting efficiency and the value being lost to residue lines.
Analyzer data can also be exported in customisable .CSV reports, or integrated with existing business intelligence software like PowerBI or Tableau.
Greyparrot Analyzer is hardware- and software-agnostic. That means we can share our data with any compatible tool, regardless of brand.
So far, examples have included robotic sorting arms, near-infrared sorters, facility management software, and more.
Learn more about Sync integrations here.
Yes. Our data security is covered by ISO27001 certification, and we also hold the UK’s Cyber Essentials certification.
Learn more about how we protect our customers' and partners' data here.
Access to the data collected in your facility is limited to the users you select, and our team. Greyparrot owns the data collected by global Analyzer units in an anonymised format, using it to continuously train and improve our AI’s waste recognition.
Analyzer gathers the same data that human samplers collect, and more.
While you can export customisable reports to suit almost any reporting requirement, it’s up to your region’s regulators whether AI data will be accepted.
We can ship Analyzers in 32 hours. For context, the industry’s average lead time is around 1 year.
Analyzer units require a simple metal mounting frame to position them above conveyor belts. We provide detailed specifications for building these frames to our customers before units arrive at their facility.
A 110-230V power supply and wired ethernet connection with 1MBps download and upload speed are also essential.
Analyzer units are designed for belts up to 1600mm wide — although solutions for wider belts up to 2800mm are available.
Every facility and scenario is different, but most of our customers complete installation within downtime they’ve already scheduled. That way, plant productivity isn’t impacted.
Analyzers don’t require much maintenance, but a bit of attention will ensure a longer lifespan. Cleaning the unit’s glass and cooling fans twice a year is usually all that’s needed.
There are two installation options for Analyzer units:
Self-installation: We designed Analyzer to be simple for our customers to quickly self-install.
Installation support: We can provide a quote for on-the-ground support with our experienced installation partners.
In either scenario, our team leads a thorough setup and calibration process once the Analyzer units are in place.
Our distribution partners cover North America and Europe, which means we can offer ongoing, on-the-ground assistance in some cases.
Learn more about our distributors and support partners here.
Instead of selling Analyzer units, we currently sell licenses, which include use of Analyzer units along with software updates and support from our cutomer success and data teams.
That allows us to give our customers access to system updates and hardware support as part of their existing subscription.
Talk to us about our pricing models here.
A Greyparrot Analyzer license includes the Analyzer hardware, along with access to the Analyzer portal for unlimited users.
We frequently update the Analyzer system to ensure it remains the most powerful AI waste analytics system available. As soon as we introduce an update to system, license-holders get instant access.
Learn more about pricing from a member of our team.
Updates include expansions to our recognition system, updates to the Analyzer portal, and the latest versions of our monitoring units.
We continuously train the AI that powers Analyzer, expanding the types of waste the system can recognise.
Our product team also introduce new ways to view, report and act on Analyzer data by adding to the Analyzer portal’s dashboards and exporting features.
The cost of an annual license varies depending on our customers’ needs, and the number of Analyzer units they deploy.
To get a quote for your recovery facility, speak to a member of our team here.
We do not currently offer free trials of the Greyparrot Analyzer system.
Reach out to us below, and a member of our team will answer your question as soon as possible.