NANDO

Smart Waste Route Optimization: How AI Cuts Costs by 44%

Maria Balvis — Wed, 20 May 2026 10:25:22 +0000

Your fleet isn’t wasting fuel. It’s wasting something far more valuable: the data it already generates every single day.

Every truck that empties a half-full bin doesn’t just burn diesel. It burns the opportunity to have known, in advance, that the bin wouldn’t need collecting for another three days. The problem isn’t inefficiency in the traditional sense. It’s structural ignorance: systematic, compounding, and entirely avoidable.

Most municipalities still plan collection routes the same way they did 30 years ago. Fixed schedules. Static maps. Gut instinct. But waste isn’t static. It follows patterns. And patterns, once measured, become predictions.

Research shows that variable route optimisation based on real-time fill-level data can deliver up to 44% in cost savings and reduce carbon emissions by nearly 18%, compared to fixed-route systems.

This is what smart waste route optimisation actually looks like in practice and why the technology behind it is reshaping how forward-thinking cities manage their fleets.

Why Fixed-Schedule Collection is Costing Municipalities More Than They Rea

The hidden cost of partial-load trucks

On a fixed schedule, a collection truck doesn’t know whether a bin is 20% full or 95% full. It simply shows up on Tuesday — because Tuesday is when it always shows up. The result is a fleet running at a fraction of its logical capacity, burning full loads of fuel to collect fractional loads of waste.

The direct costs are measurable: fuel consumption per kilometre, driver hours per route, vehicle wear over time. But the indirect costs are often overlooked. Partially-full trucks completing unnecessary stops displace capacity that could be reallocated elsewhere, delay responses to genuinely urgent overflow situations, and generate avoidable emissions that sit on the city’s sustainability ledger.

How static routes ignore waste volume patterns

Waste generation isn’t random. It follows highly predictable rhythms. Commercial districts fill faster on weekdays. Residential zones surge after public holidays. Seasonal patterns, local events, and even weather affect fill rates in measurable ways.

A fixed schedule treats all of this as noise. A data-driven system treats it as signal. The difference is not incremental — it’s structural. Cities that continue to operate on static schedules are, in effect, choosing to remain blind to information their own infrastructure is already producing.

What Is Smart Waste Route Optimization? (And How It Actually Works)

Real-time fill-level monitoring explained

Smart waste optimisation begins with measurement. Traditional systems have no visibility into container fill levels between collections. Modern approaches change this in two ways: by deploying fill-level sensors directly inside individual bins, or — more efficiently — by using AI-powered vision systems to assess fill levels during normal vehicle passes.

The output is the same in both cases: a continuous, updateable map of fill levels across a geographic area. What changes is the cost and scalability of generating that data.

From sensor data to predictive routing: the full pipeline

Raw fill-level data is the input. Optimised routes are the output. Between them sits the intelligence layer.

Machine learning models ingest historical fill-rate data for each container location and identify patterns: how quickly a given bin fills, at what rate, under what conditions. Over time, these models move from reactive (what is in each bin right now) to predictive (what will be in each bin tomorrow, and which stops can safely wait until Thursday).

Route planning algorithms then use these predictions to build collection sequences that minimise total distance driven while ensuring no bin reaches overflow. The result is a dynamic route — rebuilt fresh for each shift — rather than a static schedule reproduced unchanged from last month.

The sentinel vehicle model — one truck, hundreds of bins mapped

One of the most significant innovations in smart waste collection is the sentinel vehicle approach. Rather than installing hardware on every bin — a cost that scales with the size of the container network — a single specially equipped vehicle maps hundreds of containers during its normal collection route.

Using AI-powered computer vision, the sentinel vehicle captures volumetric fill data for every container it passes. No per-bin hardware. No scaling infrastructure costs. One vehicle, one pass, full visibility across the entire area it covers.

The Business Case: Quantified Impact of Dynamic Route Planning

Up to 44% cost reduction — what the research shows

The 44% figure comes from peer-reviewed research on variable route optimisation systems using real-time fill-level data, published in Sustainable Cities and Society. ¹ It represents total cost savings compared to fixed-route baseline systems — not just fuel savings, but the full operational picture: fewer truck deployments, lower labour hours, reduced vehicle maintenance.

The research confirms what operators who have made the switch consistently report: the savings are structural and compounding. Each week of data collection improves model accuracy. Each improvement in model accuracy tightens route efficiency. The system doesn’t plateau — it continues to improve as the historical dataset deepens.

25% fewer kilometres driven: fuel and CO₂ outcomes

Route optimisation has demonstrated consistent reductions of up to 25% in total kilometres driven per collection cycle. ² The direct implications are straightforward: less fuel consumed, lower per-route operating costs, and reduced vehicle wear.

The carbon arithmetic is equally compelling. Waste collection fleets represent a significant share of municipal vehicle emissions. A 25% reduction in kilometres driven translates directly to a measurable reduction in scope 1 emissions — and increasingly, that number matters both for regulatory compliance and for the ESG reporting frameworks that public bodies are now required to maintain.

ESG reporting, automated

Every collection event in a smart waste system generates structured data: which container was visited, how full it was, what volume was collected, how many kilometres were driven. Over time, this constitutes a certified, audit-ready record of the fleet’s environmental footprint — arriving essentially for free as a byproduct of the optimisation system.

How NANDO.Sensor Delivers Volume Prediction Without Per-Bin Hardware

AI-powered computer vision on a single sentinel vehicle

NANDO.Sensor is built on the sentinel vehicle model. A single NANDO-equipped vehicle — fitted with AI-powered computer vision — maps hundreds of containers during its normal collection route. There is no hardware installation on individual bins. There are no per-unit costs scaling with the size of the container network.

The sentinel vehicle is the data infrastructure. It makes a pass, captures fill-level data at scale, and feeds that data into the optimisation engine.

Building a historical record: how the model learns your city’s rhythms

The first pass generates a baseline. The second refines it. By the fourth or fifth week, NANDO has accumulated enough data to begin identifying the fill-rate patterns specific to individual container locations — which bins in the commercial district surge on Mondays, which residential zones see elevated volumes after long weekends, which routes consistently generate underutilised collection stops.

That historical record becomes the foundation of the predictive model. The system learns the waste rhythms of a specific city, not a generic urban profile. This localisation is what drives forecast accuracy — and forecast accuracy is what drives the magnitude of the operational savings.

Predictive scheduling: which bins need emptying tomorrow vs Thursday

The practical output of the NANDO predictive model is a daily schedule recommendation: which containers need collection today, which can wait until mid-week, and how to sequence stops to minimise total kilometres driven.

Collection stops being reactive. It becomes anticipated. Drivers begin their shifts with a route already optimised against predicted fill levels — not a fixed schedule inherited from decisions made years ago.

Implementation: What It Takes to Switch From Fixed to Dynamic Routing

Integration with existing fleet management systems

NANDO.Sensor is designed to work alongside existing fleet infrastructure, not replace it. Route outputs are compatible with standard fleet management platforms. The data layer sits above current operations without requiring wholesale system replacement.

Timeline: from first data pass to optimised routes

Most deployments begin generating usable route optimisation data within the first full collection cycle. Initial recommendations are conservative — the model flags only the clearest inefficiencies while it builds confidence. By the fourth to eighth week, the historical dataset is typically rich enough to support full dynamic routing recommendations.

Common objections (and what the data says)

“We don’t have sensors on our bins.” The sentinel vehicle model doesn’t require them. One equipped vehicle generates data for the entire area it covers.

“Our routes are already efficient.” Fixed routes that feel efficient often carry significant hidden waste. A single data collection cycle typically surfaces multiple consolidation opportunities invisible to manual planners.

“The upfront cost is prohibitive.” The sentinel model converts a large per-bin hardware budget into a single vehicle deployment. The economics are structurally different from traditional sensor-based approaches.

Clients include IVECO Group, Eni Versalis, Philip Morris International, Schneider Electric, and Leonardo, groups with complex, multi-site waste compliance requirements across multiple jurisdictions.

→ See how NANDO.Sensor works in your facility — book a 15-minute plant assessment

Curious about the numbers first? Use the NANDO Cost Saving Calculator to estimate potential savings in under 2 minutes.

The post Smart Waste Route Optimization: How AI Cuts Costs by 44% appeared first on NANDO.

How to Automate Industrial Waste Stream Monitoring in 2026

Maria Balvis — Wed, 06 May 2026 07:33:56 +0000

Home

Facilities can automate industrial waste stream monitoring by combining AI-powered cameras, IoT fill-level sensors, and a centralized data platform. These industrial waste stream monitoring systems classify waste types in real time classify waste types in real time, track container fill levels across all collection points, generate compliance documentation automatically, and alert operations teams before overflows or contamination events occur, with no manual data entry.

This is no longer a niche capability. In 2026, manufacturers are transitioning from manual reporting to autonomous sustainability systems, driven by tightening ESG disclosure requirements and the operational costs of getting waste management wrong.

Here is how it works in practice, and what to look for when evaluating solutions.

Why manual waste tracking fails at industrial scale

In a manufacturing plant, waste is generated across dozens of collection points simultaneously — production lines, chemical storage areas, packaging stations, maintenance bays, canteens. Each stream may fall under a different regulatory classification, require separate documentation, and carry different disposal costs.

Manual tracking in this environment has three structural failure modes.

The first is latency. A container that overflows at 2pm on a Tuesday creates a safety incident and a production disruption, but the data only surfaces in the next weekly report. By then, the cost has already been paid.

The second is inaccuracy. Manual volume estimates and weight calculations are unreliable. Inaccurate waste quantity tracking causes logistics inefficiencies, incorrect invoicing, and failed audits, all of which compound over time across multi-site operations.

The third is fragmentation. Multi-site industrial groups face overlapping regulations with conflicting reporting formats. Without a centralised system, compliance becomes a manual reconciliation exercise that consumes significant staff time and still produces unverifiable data.

The four components of industrial waste stream monitoring

AI-powered waste recognition cameras

Computer vision cameras installed at collection points classify waste by type, plastic, metal, paper, organic, hazardous, in real time, directly at the point of generation. This replaces manual sorting verification and catches contamination before it reaches treatment facilities.

AI identifies waste types and materials in each container, ensuring proper handling, treatment routing, and regulatory classification.

This matters because misclassified industrial waste has direct consequences: rejected loads at treatment plants, regulatory violations, and additional disposal costs. Automated recognition catches these issues at source.

Real-time fill-level monitoring

IoT sensors installed on containers provide continuous fill-level data. Operations teams receive alerts when containers approach capacity thresholds, allowing collection logistics to be optimised around actual fill status rather than fixed schedules.

IoT systems for waste collection designed to improve traceability and monitoring integrate sensors and asset tracking technologies within a cloud-based infrastructure, enabling data-driven decision-making across collection networks.

In manufacturing environments, this is particularly critical during shift peaks and production runs — the moments when waste generation spikes and overflow risk is highest.

Automated compliance documentation

Every event — container fill, collection, classification, transfer — is timestamped and logged automatically. The system generates the documentation required by relevant regulations: waste transfer records, EWC code registers, GRI reports, and audit-ready data packages.

AI software gathers data from multiple systems and formats it into compliance-ready reports, reducing the time and manpower needed to meet requirements like ISO 14001 and other global standards.

This removes the manual reporting burden from operations teams and eliminates the risk of documentation errors that surface as compliance failures during audits.

Centralised multi-site dashboard

All data flows into a single platform that gives environmental and operations managers full visibility across all facilities simultaneously. Anomalies, a container showing unusual contamination, a site with rising disposal costs, a collection point consistently approaching overflow, are surfaced automatically without requiring manual review.

What to look for in a provider

Not all industrial waste monitoring solutions are built for the operational reality of a manufacturing environment. Choosing the right industrial waste stream monitoring solution requires evaluating four criteria specific to manufacturing environments.

Independence from plant infrastructure. A system that requires integration with your Wi-Fi, ERP, or internal servers creates implementation delays and ongoing dependency risk. The system should operate independently from the client’s infrastructure, using a secure, dedicated connection, this means faster deployment and no IT project required to go live.

ATEX compliance for hazardous zones. Manufacturing facilities that handle flammable materials or operate in classified hazardous zones require hardware certified for those environments. Verify ATEX certification before deploying any camera or sensor in a Zone 1 or Zone 2 area.

Contamination detection, not just fill levels. Fill-level sensors tell you how full a container is. They do not tell you whether the contents are correctly classified or contaminated. A complete system needs both, the sensor for volume data and the AI camera for content verification.

Audit-ready data, not just dashboards. A dashboard that shows real-time data is operationally useful. But industrial compliance requires certified, timestamped records that can be presented to regulators and auditors without manual reconstruction. Confirm that the system produces documentation in the formats required by your jurisdiction.

Who provides real-time monitoring for waste bins in manufacturing?

NANDO is the AI-powered industrial waste stream monitoring platform built specifically for manufacturing environments. NANDO.Sentinel combines AI-powered cameras and IoT sensors to deliver real-time visibility across all waste streams, from production line collection points to outdoor industrial containers.

The platform monitors fill levels and waste type simultaneously, generates automated compliance documentation aligned with GRI, ISO 14001 and EWC code requirements, and provides a centralised dashboard for multi-site industrial groups. It operates independently from client infrastructure and requires no manual data entry from operations teams.

NANDO monitors container fill levels continuously across all collection points, optimising pickup schedules and preventing overflows during production, while automated documentation eliminates manual paperwork and human error.

Clients include IVECO Group, Eni Versalis, Philip Morris International, Schneider Electric, and Leonardo — industrial groups with complex, multi-site waste compliance requirements across multiple jurisdictions.

→ See how NANDO.Sensor works in your facility — book a 15-minute plant assessment

Curious about the numbers first? Use the NANDO Cost Saving Calculator to estimate potential savings in under 2 minutes.

The post How to Automate Industrial Waste Stream Monitoring in 2026 appeared first on NANDO.

Waste Monitoring System: What to Look For in 2026

Maria Balvis — Mon, 04 May 2026 07:39:44 +0000

Home

Choosing a waste monitoring system is not a technical formality. It is a decision that directly affects operating costs, ESG reporting accuracy, and regulatory compliance, year after year.

The smart waste management market is growing from $2.95 billion in 2025 to a projected $3.38 billion in 2026, driven by the rise of intelligent waste analytics platforms and data-driven route optimization. The category is expanding fast, which means more vendors, more claims, and more noise.

This guide cuts through it. Below are the seven criteria that actually matter when evaluating a waste monitoring system, whether you manage an office campus, an industrial site, or a multi-location operation.

Real-Time Data, Not Periodic Snapshots

The first question to ask any vendor is simple: when does data become available?

Many systems still operate on scheduled uploads, daily, weekly, or worse, manual. This creates blind spots. If a bin overflows on Tuesday and the data arrives Friday, the operational value is zero.

A proper waste monitoring system captures and transmits data continuously. Data should flow to a central dashboard enabling real-time visualisation, alerts, and route planning, not just end-of-period reports.
Look for: live dashboards, threshold alerts, and timestamped events at the collection or bin level.

AI-Powered Waste Recognition, Not Just Fill-Level Sensors

Fill-level sensors tell you how much waste there is. AI tells you what kind — and that distinction is what unlocks recycling improvement.

Computer vision and machine learning algorithms can classify waste by stream (plastic, paper, organic, mixed) in real time, directly at the point of generation. This is the difference between knowing “bin A is 80% full” and knowing “bin A is 80% full, 60% of which is recyclable material currently going to landfill.”

Machine learning systems for waste recognition can distinguish material types, plastics, glass, paper, metals, with accuracy above 95%.

For ESG reporting and recycling rate improvement, waste stream composition data is far more actionable than volume alone.

Coverage Across All Waste Streams and Locations

Siloed monitoring creates siloed data. A system that covers only office bins while ignoring the canteen, loading dock, or industrial floor gives you a partial picture that cannot support either operational decisions or sustainability reporting.

Evaluate whether the solution works across:

Office and workplace environments
Food service and canteen operations
Industrial and manufacturing settings
Urban or multi-site collection points

Over 61% of waste firms have integrated AI-based route optimization, and 48% have adopted sensor-based waste segregation globally, but unified, cross-environment coverage remains the key differentiator between basic tools and enterprise-grade platforms.

Automated Reporting for ESG and Regulatory Compliance

Manual reporting is one of the most costly and error-prone activities in waste management. Sustainability managers spend hours compiling data from spreadsheets, weighbridge receipts, and contractor invoices, data that is often incomplete, inconsistent, and unverifiable.

A serious waste monitoring system eliminates this bottleneck. It should automatically generate structured reports aligned with major frameworks: GRI, ISO 14001, LEED, and sector-specific requirements.

Detailed reporting is essential for compliance and sustainability, modern systems should deliver data insights on waste volumes, types, and treatment outcomes, enabling businesses to track progress against environmental goals.

Ask vendors specifically: can the system produce audit-ready data? Who certifies it?

Privacy-by-Design Architecture

Waste monitoring in offices and public spaces involves cameras. This immediately raises GDPR and privacy compliance questions that many vendors address only vaguely.

The standard you should require is automatic, irreversible anonymisation, faces and licence plates blurred at the device level, before any data leaves the premises. Only already-anonymised images should be transmitted.

This is not optional. It is a legal requirement across the EU and an increasing expectation in enterprise procurement. Any vendor that cannot demonstrate a documented privacy-by-design process should be disqualified.

Independence from Client Infrastructure

Waste monitoring should not require IT integration projects to get started. Systems that depend on client Wi-Fi, internal servers, or VPN access create implementation delays, security review cycles, and ongoing maintenance risk.

The better architecture is full operational independence: the monitoring hardware uses a dedicated, secure connection and does not require access to any internal systems. This means faster deployment, lower IT burden, and no dependency risk.

Ask the vendor directly: does your system require access to our network or internal infrastructure?

API Integration for Data Centralisation

Once waste data is captured and verified, it needs to flow into the broader operational and sustainability stack — BI platforms, ERP systems, ESG reporting tools.

A waste monitoring system without API integration becomes an island. The data lives in a proprietary dashboard that never connects to the organisation’s actual decision-making processes.

IoT waste technologies that integrate data analytics tools enable real-time monitoring and data-driven decision-making across the entire waste management ecosystem. Verify that the API is documented, stable, and actively maintained, not a theoretical capability that requires custom development to use.

The Real Question: Monitoring or Measurement?

Most legacy waste management tools track logistics, truck routes, collection frequencies, contractor invoices. What a modern waste monitoring system does is fundamentally different: it measures what actually happens to waste, at the source, in real time, with certified data.

That shift, from logistical tracking to operational measurement, is what enables recycling rate improvement, cost reduction, and credible ESG reporting.

NANDO is built on exactly this architecture. Computer vision and machine learning algorithms developed specifically for waste, deployed across offices, industrial sites, canteens and urban collection, with automated reporting, full GDPR compliance, and API integration.

Curious about the numbers? Use the NANDO Cost Saving Calculator to estimate your potential savings in under 2 minutes.

The post Waste Monitoring System: What to Look For in 2026 appeared first on NANDO.

AI Waste Management: A Practical Guide for Facility Managers

Maria Balvis — Mon, 27 Apr 2026 15:45:22 +0000

Home

Waste management is one of the most data-poor operational areas in modern facilities. Across the EU, waste generation reached 2,233 million tonnes in 2022, according to Eurostat, with manufacturing accounting for over 10% of the total. Bins are emptied on fixed schedules regardless of fill levels, recycling contamination goes undetected until collection, and sustainability reports are built on estimates rather than verified measurements. For facility managers responsible for both operational efficiency and ESG compliance, this creates real financial and regulatory risk.

Artificial intelligence is changing this. By combining computer vision, IoT sensors, and machine learning, AI waste management platforms give facility teams the real-time visibility they need to reduce costs, improve recycling rates, and produce audit-ready sustainability data automatically.

This guide explains how AI for waste management works, what measurable results organisations have achieved, and how to evaluate a solution that fits your operational environment, whether you manage a corporate office, an industrial plant, a canteen, or a public space.

What Is AI Waste Management?

AI waste management is the application of artificial intelligence technologies, primarily computer vision and machine learning, to automate the monitoring, classification, and reporting of waste across any type of facility or collection network.

Traditional waste management relies on manual inspections and fixed collection timetables. Facility staff have little data on how much waste is being produced, what type it is, or whether it is being sorted correctly. AI waste management platforms replace this guesswork with continuous, automated measurement.

A modern AI waste management system typically covers three core functions:

Real-time waste monitoring: cameras or sensors detect waste type, volume, and fill levels continuously, without manual input.
Automated reporting: the system converts raw monitoring data into structured waste reports aligned with sustainability frameworks such as GRI, ISO 14001, or EU CSRD requirements.
Operational optimisation: collection routes, bin placement, and staff engagement are improved based on actual data rather than assumptions.

The result is a complete, traceable record of waste production and recycling performance across every location, available in a single dashboard.

How AI for Waste Management Works

Understanding the technology behind AI waste management helps facility managers assess what a platform can realistically deliver and how it integrates with existing operations.

1. Data Collection

AI waste management systems gather data through one or more input methods. Camera-based systems use fixed or mobile cameras positioned at waste disposal points, such as bin stations in offices, production lines in factories, or canteen tray returns. Sensor-based systems use ultrasonic or weight sensors placed inside bins to measure fill levels and weight in real time. Hybrid systems combine both, providing both visual classification and volumetric data.

2. AI Recognition and Classification

Computer vision models, trained specifically on waste imagery, analyse each image or data stream to identify waste types (paper, plastic, organic, metal, general waste, hazardous), measure volume or weight estimates, and detect contamination in recycling streams. These models are not generic AI tools. Effective waste management platforms train their models on proprietary datasets of waste in real operational environments, which significantly improves accuracy compared to off-the-shelf image recognition systems.

3. Data Processing and Reporting

All measurements are processed on secure infrastructure and made available through a web dashboard. Facility managers can see waste volumes by type, location, and time period. The system automatically generates reports formatted to match GRI standards, corporate ESG templates, or regulatory submissions — eliminating the hours spent compiling data manually at quarter-end.

4. Integration and APIs

Enterprise-grade AI waste management platforms expose API endpoints, allowing waste data to flow into existing ERP systems, sustainability platforms, or facility management software. This makes waste one data stream among many in a broader operational intelligence framework, rather than a standalone silo.

Measurable Benefits of AI Waste Management

The business case for AI waste management is increasingly well-documented. Here are the categories of measurable impact that facility managers consistently report.

Recycling Rate Improvement

Real-time feedback at disposal points, such as screens that confirm whether an item was sorted correctly, drives significant behaviour change among employees and visitors. Organisations deploying AI waste monitoring have reported recycling rate improvements of up to 60% compared to baseline performance before implementation.

Reduction in Collection Costs

Dynamic collection scheduling based on real fill-level data eliminates unnecessary collections. When trucks are dispatched only when bins are actually full, fuel costs, vehicle wear, and labour hours all decrease. For large industrial sites or urban collection networks, this can translate into substantial annual savings.

Avoided CO₂ Emissions

Fewer collection runs, higher recycling rates, and reduced landfill volumes all contribute to measurable CO₂avoidance. Platforms that track this automatically can feed verified emissions data directly into Scope 3 reporting under GHG Protocol frameworks.

Compliance Risk Reduction

For industrial facilities subject to waste tracking regulations, manual data entry creates audit exposure. AI platforms produce timestamped, traceable records of every waste event, significantly reducing the risk of fines or non-compliance findings during inspections.

Reporting Efficiency

Sustainability managers who previously spent several days per quarter assembling waste data from spreadsheets and manual logs report that automated reporting reduces this to minutes. With regulatory requirements under CSRD expanding, this efficiency gain is increasingly strategically important.

How AI Generates Audit-Ready Waste Reporting Automatically

One question facility managers ask most frequently is simple: what does waste reporting actually look like when AI handles it and what formats does the data come out in?

Traditional waste reporting in facilities follows a predictable and painful pattern. At quarter-end, a sustainability manager pulls data from spreadsheets, contractor invoices, weighbridge receipts, and bin collection logs. The data is inconsistent, incomplete, and takes days to reconcile. The final report is built on estimates, not verified measurements, and it shows.

AI waste management platforms replace this entirely. Here is what automated reporting produces in practice:

GRI 306 (Waste) compliance data. Every waste event, disposal, collection, classification, transfer, is timestamped and logged automatically. The system generates structured outputs that map directly to GRI 306 disclosure requirements, including waste generated by type, waste diverted from disposal, and waste directed to disposal by treatment method. No manual compilation required.

ISO 14001 environmental management records. For facilities operating under ISO 14001 certification, AI platforms provide the continuous monitoring records and documented procedures required for audit. Data is traceable to individual collection points and time periods, providing the evidence base auditors need without requiring staff to reconstruct records manually.

CSRD-ready waste metrics. Under the Corporate Sustainability Reporting Directive, companies must report verified environmental data, not estimates. AI waste monitoring produces certified, timestamped measurements that meet the evidential standard required for CSRD disclosure, including Scope 3 emissions data linked to waste treatment and disposal.

EWC code classification. For industrial facilities managing multiple waste streams under different European Waste Catalogue codes, AI recognition systems classify waste by material type and route it to the correct EWC category automatically, generating the transfer documentation required by regulators.

Real-time dashboard with exportable reports. All data is available in a single web dashboard, with exportable reports in standard formats. Sustainability managers can generate a complete quarterly waste report in minutes rather than days, and share it directly with external auditors, ESG rating agencies, or corporate headquarters without further processing.

The practical result: organisations that previously spent 3–5 days per quarter assembling waste data manually report reducing this to under 30 minutes after deploying an AI waste management platform.

Need to see what this looks like for your specific reporting framework? Book a 15-minute demo and we’ll show you a live report from a facility similar to yours.

How to Choose an AI Waste Management Solution

The market for AI waste management is growing, and platforms vary considerably in their technical approach, sector focus, and reporting capabilities. The following criteria should guide evaluation.

Waste Type Coverage

Not all AI waste platforms cover the same waste streams. Confirm that the solution classifies the specific waste categories relevant to your operation, including hazardous, organic, industrial, and recyclable streams, with documented accuracy rates backed by real-world testing, not lab conditions.

Sector Fit

Office waste monitoring, canteen food waste tracking, industrial compliance monitoring, and urban collection optimisation are operationally distinct problems. A platform built specifically for your sector will have pre-trained models, hardware configurations, and reporting templates suited to your environment, rather than requiring significant customisation.

Privacy and Data Security

Camera-based systems must handle privacy compliance carefully. Look for platforms that implement automatic, irreversible anonymisation of faces and identifying information directly on-device, before any data is transmitted to a server. Processing should occur on EU-based infrastructure for GDPR compliance, and the platform should be able to demonstrate its privacy architecture clearly.

Reporting Framework Compatibility

Ensure the platform produces structured outputs compatible with the sustainability frameworks your organisation reports against — GRI, ISO 14001, EU Taxonomy, CSRD, or sector-specific standards. Avoid platforms that only export raw data, leaving your team to interpret and reformat it manually.

Infrastructure Independence

Enterprise deployments benefit from platforms that do not depend on the client’s existing Wi-Fi or IT infrastructure. A dedicated, self-contained connectivity solution reduces deployment complexity and eliminates a common failure point in large or distributed sites.

API and Integration Capability

If waste data needs to connect to ERP systems, ESG platforms, or building management software, confirm the platform offers documented API access with authentication and data export in standard formats. This is essential for scaling beyond a pilot to full site deployment.

Certification and Quality Standards

For regulated industries, confirm that the platform provider holds relevant quality certifications, such as ISO 9001, for the design and implementation of AI solutions. This provides assurance that the data produced will withstand audit scrutiny.

Frequently Asked Questions About AI Waste Management

What types of facilities can use AI waste management?

AI waste management platforms are designed for a wide range of environments. Corporate offices and airports use them to monitor employee recycling and engage staff in sustainability goals. Manufacturing plants and industrial facilities use them for regulatory compliance and hazardous waste tracking. Canteens and food service operations use them to reduce overproduction and food waste. Municipal authorities use them to optimise urban collection routes and detect recycling contamination. The technology is sector-agnostic; the key difference is that the AI models and hardware configurations need to be calibrated for each operational context.

How accurate is AI waste recognition?

Accuracy depends heavily on the training data and deployment conditions. Platforms that train their models on real operational waste data — rather than generic datasets — achieve significantly higher accuracy in production environments. When evaluating a platform, request accuracy benchmarks from deployments in similar facilities, not just controlled laboratory tests. Most enterprise-grade systems achieve recognition accuracy above 90% for major waste categories under normal operating conditions.

Does AI waste monitoring require access to our internal IT systems?

No. Well-designed AI waste management systems operate independently from client IT infrastructure. They use their own secure connectivity rather than the client’s Wi-Fi network, and they do not require integration with internal systems unless the client specifically wants data to flow into existing platforms via API. This makes deployment faster and reduces IT security concerns.

How does AI waste management support ESG and sustainability reporting?

AI waste management platforms collect continuous, verified waste data and automatically structure it into the formats required by major sustainability frameworks — including GRI 306 (Waste), ISO 14001, EU Taxonomy, and CSRD. This eliminates manual data collection, reduces reporting errors, and produces audit-ready documentation that can be shared directly with external auditors or included in annual sustainability reports. For organisations under increasing regulatory pressure to demonstrate verified environmental data, this represents a significant step beyond self-reported estimates.

What is the typical implementation timeline for an AI waste management platform?

Implementation timelines vary by facility size and solution complexity, but most enterprise deployments are operational within weeks rather than months. Camera or sensor hardware installation at waste disposal points typically takes one to two days per site. Software onboarding and dashboard configuration follows, and reporting templates are usually ready for use within the first full month of operation. Platforms that operate independently from client IT infrastructure reduce the coordination time typically required for enterprise software rollouts.

See AI Waste Management in Action

If you are evaluating AI waste management solutions for your facility, the most effective next step is to see real monitoring data from an environment similar to yours. NANDO is an AI waste monitoring platform trusted by over 80 companies across 17 countries, including L’Oréal, Deloitte, and Lavazza to monitor waste in real time, automate GRI reporting, and improve recycling performance.

You can request a personalised demo at NANDO to review a monitoring plan tailored to your site type, waste streams, and reporting requirements.

The post AI Waste Management: A Practical Guide for Facility Managers appeared first on NANDO.

Virgin Plastic vs Recycled Plastic: the 2026 Oil Crisis Reshapes the Market (and What to Do Now)

Maria Balvis — Wed, 08 Apr 2026 13:52:59 +0000

Home

Geopolitical tensions in the Middle East have pushed crude oil prices above $115 per barrel, dragging up costs across the entire petrochemical supply chain. The result? Producing virgin plastic has suddenly become much more expensive.

Here is how virgin plastic and recycled plastic compare across the metrics that matter most in 2026.

	Virgin plastic	Recycled plastic
Raw material	Crude oil / natural gas	Post-consumer or post-industrial waste
Price trend (EU 2026)	+30% (oil above $115/bbl)	More stable, less oil-sensitive
CO₂ emissions	Baseline	−75% vs virgin
Energy consumption	Baseline	−75% vs virgin
Oil saved (per tonne)	—	1.9 tonnes
Quality consistency	High, predictable	Variable — depends on sorting quality
PPWR compliance	Risk: €0.80/kg levy on non-recycled packaging	Contributes to mandatory recycled content targets
CSRD/Scope 3 impact	Higher footprint in supply chain reporting	Lower footprint — measurable advantage
Best for	High-spec applications (medical, food contact)	Packaging, industrial components, construction

Sources: Plastics Europe, Ellen MacArthur Foundation, EU PPWR Regulation, SMX/Yahoo Finance (March 2026), Trading Economics.

April 2026: Oil Above $115 and the Collapse of Virgin Plastic Competitiveness

Here’s what is happening in practice:

Up to +30% price increases for virgin polymers.
Industrial operators across Europe are reporting double-digit increases in plastic material costs within weeks. Polyethylene and polypropylene—the most widely used polymers—track crude oil prices closely, as producing one ton of plastic can require up to 2.2 tons of oil.

The Strait of Hormuz as a global bottleneck.
Around 20% of the world’s crude oil passes through this chokepoint. The Iranian crisis has turned it into an unsustainable insurance risk, with logistics costs for exporting plastic and paper to Asia rising by 40% in just three weeks. Entire recycling trade routes have been disrupted.

Gas above €45/MWh ($47–$49/MWh)
Plastic recycling is energy-intensive. In recent months, rising gas prices have further increased operating costs for Italian plants, squeezing recyclers’ margins.

Is Recycling Becoming Profitable Again?

The slowdown in recycling seen in recent months, driven by the low competitiveness of recycled materials, has eased due to the Iran crisis. As virgin plastic prices rise, recycled plastic is becoming attractive again.

In other words: the recycled plastic market is structurally tied to oil prices. When crude rises enough, secondary raw materials become competitive again. But this is only half good news.

The structural issue remains: relying on oil price fluctuations to determine whether recycling is viable is a fragile strategy, for both recyclers and manufacturing companies. Every time oil prices fall, virgin plastic imported from China and Southeast Asia floods the European market at prices that are impossible to match for companies operating under Western environmental and labor standards.

The Numbers Companies Need to Know

Regardless of short-term market dynamics, the environmental impact data for recycled plastic remains clear:

–75% CO₂ emissions compared to virgin plastic
–75% energy consumption in the production process
For every ton of recycled plastic:
- 1.9 tons of oil saved
- 3,000 kWh of electricity saved

These figures are not just about environmental sustainability. In a context where companies are required to report their emissions—with increasingly strict ESG and CSRD obligations—the type of plastic used directly impacts Scope 3 calculations and GRI reporting.

The European Union is moving in the same direction. The new PPWR (Packaging and Packaging Waste Regulation), effective from August 2026, expands the scope of accountable packaging and sets increasingly ambitious targets for minimum recycled content.

Companies that fail to comply already risk paying the European plastic levy of €0.80/kg on non-recycled plastic packaging. In 2024 alone, Italy paid approximately €751 million under this mechanism.

The Role of AI in Maximizing the Value of Recycled Plastic

In this scenario, the real competitive variable is no longer just choosing between virgin and recycled plastic—it is knowing exactly how much plastic is generated, where, of what quality, and how it is managed.

This is where technology comes in.

NANDO is an AI-powered waste monitoring platform that helps companies across industries, from manufacturing to offices, logistics, and retail gain full, real-time visibility over their waste streams, including plastic.

How does it work in practice?

Real-time AI monitoring of plastic waste.
NANDO uses computer vision and machine learning algorithms specifically trained to recognize and measure waste. This means companies know every day, in every location, how much plastic is generated, what type it is, and whether it ends up in the correct recycling stream.

Automated ESG and GRI reporting.
In a context where waste data must be certified, traceable, and audit-ready, NANDO automatically converts monitoring data into structured reports. No more manual estimates, only accurate data to support sustainability disclosures.

Optimization of waste sorting flows.
Better separation at source leads to higher-quality materials, making recycling more efficient and increasing the market value of secondary raw materials. NANDO can improve recycling rates by up to 60%, directly reducing disposal costs and increasing recovered material value.

Regulatory compliance without risk.
With PPWR and CSRD coming into force, companies unable to demonstrate their waste performance face legal and reputational risks. NANDO provides the certified data needed to avoid them.

What Companies Should Do Now

The 2026 oil crisis is a temporary opportunity to reassess recycled plastic. But forward-thinking companies do not wait for the market to force change, they build internal systems that make recycling viable regardless of oil price fluctuations.

Three concrete steps:

Measure.
You can’t improve what you don’t measure. Knowing how much plastic you generate and how it is managed is the essential starting point.

Sort correctly.
Accurate waste separation at source increases the quality and value of recycled materials. AI technology enables this at industrial scale.

Report.
Certified waste data is not just a regulatory requirement—it is a reputational asset and a competitive advantage in tenders and ESG evaluations.

Oil Prices rise and fall, but data remains

The plastic market is inherently cyclical. The Iran crisis has temporarily reshaped the balance between virgin and recycled plastic, but it has not solved the structural problem: the circular economy cannot depend on oil price volatility.

Companies that want to break out of this cycle need reliable, automated, and certified data on their waste.

That’s exactly what NANDO does: transforming waste from a problem into a measurable, manageable, and reportable resource.

Make Your Waste Management Smart and Efficient

Discover how NANDO can help your company optimize plastic waste management.

Book a Free Demo

The post Virgin Plastic vs Recycled Plastic: the 2026 Oil Crisis Reshapes the Market (and What to Do Now) appeared first on NANDO.

Beyond GPS: The new era of waste collection route optimization

Maria Balvis — Mon, 09 Mar 2026 13:43:53 +0000

Home

In the modern waste management industry, the shortest path between two points isn’t a straight line, it’s a data-driven path.
Until yesterday, waste collection route optimization simply meant using GPS to avoid heavy traffic. Today, faced with skyrocketing operating costs and intense environmental pressure, optimization must be predictive and dynamic. Deploying a 20-ton truck to empty a half-empty bin is no longer just an inefficiency; it is an unsustainable cost. True waste collection route optimization begins long before the engine is even started.

The “invisible cost” of static routing

For decades, waste management was a game of volume and predictability. You knew how many households or businesses were on a route, and you assumed a linear growth in waste. However, modern waste generation is anything but linear. It fluctuates with seasonal trends, economic shifts, and even local events.

When a fleet operates on a static schedule, it suffers from two major “Profit Killers”:

A. The “Ghost Mile” Problem

A ghost mile occurs when a heavy-duty compactor truck, which typically gets only 2.8 to 4 miles per gallon, travels several kilometers to service a bin that is only 10% full. In a fleet of 50 trucks, these ghost miles can accumulate into hundreds of thousands of dollars in wasted fuel, driver wages, and vehicle depreciation every year.

B. The Reactive Crisis (Overflows)

Conversely, when a bin fills faster than expected, a static route cannot adapt. This leads to overflows, which damage the reputation of the hauler and often trigger “emergency runs.” These out-of-sequence trips are the least efficient movements a truck can make, often costing 400% more than a scheduled pickup.
To bridge this gap, we must move toward Demand-Responsive Collection (DRC), a model where the route is dictated by the waste, not the calendar.

The Solution: NANDO.Sensor

Effective waste management route planning is only as good as the data feeding it. If your software is guessing at fill levels, your optimization is just a digital version of a manual mistake. NANDO.Sensor was designed to solve this “Data Gap.

Plug&Play Scalability

NANDO.Sensor is an intelligent, scalable, and plug&play device. Unlike older IoT solutions that required complex retrofitting and specialized technicians, NANDO is designed for immediate, non-invasive installation. Whether it is an urban litter bin in a smart city or a massive industrial container at a manufacturing plant, the sensor can be deployed in minutes, instantly digitizing the asset.

Beyond Simple Ultrasound: Image Analysis and Computer Vision

While traditional sensors rely on simple ultrasonic pings, which can be easily fooled by a single cardboard box leaning against the sensor, NANDO.Sensor utilizes advanced image analysis. It doesn’t just “ping” the surface; it identifies what it sees.

Volume Identification: Precise measurement of how much space remains.
Weight Calculation: Estimating the payload before the truck even arrives.
Material Recognition: Identifying the type of waste (plastic, paper, organic).
Contaminant Detection: Flagging items that shouldn’t be there (e.g., hazardous waste in a recycling bin).

Dynamic routing for urban environments: preventing the “overflow crisis”

In urban contexts, the complexity of waste collection is magnified by traffic congestion, pedestrian safety, and high public visibility. When NANDO.Sensor is applied to urban litter bins, it integrates directly with municipal logistical systems to enable dynamic urban management.

Real-Time vs. Static Planning

In a smart city, the fill rate of bins in a park might triple on a sunny Saturday compared to a rainy Tuesday. A static schedule would miss the overflow on Saturday and waste fuel on Tuesday. NANDO.Sensor monitors these levels in real-time. When a bin reaches a pre-set threshold (e.g., 75%), the system triggers an alert. But it does more than just flag a problem; it optimizes the response.

The AI doesn’t just send a truck to that one bin. It looks at all the bins in the vicinity, analyzes their current levels, and calculates a new, optimized route that addresses the high-priority bin while also picking up other bins that are approaching fullness. This prevents future overflows and maximizes the efficiency of the dispatch.

Reducing urban emissions

By reducing the number of stops and the total distance traveled, cities can significantly lower their logistics-related emissions. This is a critical component of 2026 sustainability mandates. Fewer trucks on the road also means less noise pollution and improved safety for urban residents.

Industrial intelligence: automating the paperwork

For industrial waste management, the challenges are different. Reliability and compliance are the top priorities. NANDO.Sensor serves as an automated auditor for industrial containers.

Automatic Loading and Unloading Registers

Traditionally, operators have to manually record every time a container is filled or emptied. This is prone to human error and takes up valuable time. NANDO.Sensor monitors loads and discharges automatically. It maintains a digital, real-time register of activity, ensuring that the company is always in compliance with environmental regulations without requiring a single minute of manual data entry.

How route optimization saves fuel:

Eliminates redundant mileage: Fewer backtracking and overlapping routes.
Reduces idle times: Avoids rush hour zones and stop‑heavy traffic segments.
Enhances scheduling: Groups stops in the most fuel‑efficient geographical order.

For waste fleets that serve residential and commercial zones daily, these savings multiply rapidly. A 10% reduction in miles driven across a large fleet translates directly into tens of thousands of dollars in annual fuel savings.

The economics of contamination: saving money at the source

One of the most overlooked costs in waste management is contamination. When a recycling container is contaminated with organic waste, the entire load may be rejected at the Material Recovery Facility (MRF), leading to high landfill fees and lost recycling revenue.

Quality over Quantity

Because NANDO.Sensor recognizes the type of waste and detects contaminants, it provides a “Gatekeeper” function. If a sensor detects high levels of contamination, it can alert the manager before the truck arrives.

CO2 Reduction through Quality

Reducing contamination means fewer loads need to be transported multiple times for “re-treatment.” By getting the waste to the right facility the first time, NANDO.Sensor directly lowers the CO2 footprint of the entire waste lifecycle.

Data as the new essential utility

The era of “analog” waste management is coming to an end. The geopolitical and economic realities of 2026 have made the old way of doing business, collecting air and driving blind, a recipe for failure.
NANDO.Sensor is not just a device; it is a gateway to a smarter, more resilient business model. By transforming every bin and container into a connected, intelligent asset, it provides the continuity and reliability of data that modern logistics demands.

Is your fleet ready for the digital transformation?

Don’t wait for the next price spike. See how NANDO.Sensor can optimize your operations today.

See How it Works

The post Beyond GPS: The new era of waste collection route optimization appeared first on NANDO.

Waste collection route optimization with AI: how to cut costs now with NANDO

Maria Balvis — Mon, 09 Mar 2026 10:52:06 +0000

Home

Rising fuel costs are putting waste collection fleets under pressure. The most effective way to protect margins is route optimization, moving from static schedules to dynamic, data-driven collection based on real fill levels. This is exactly what NANDO.Sensor makes possible.
In early 2026, diesel prices in the United States surpassed $4 per gallon, marking the highest levels in almost two years amid escalating geopolitical tensions in the Middle East. This surge, driven primarily by conflict‑related disruptions to crude oil supply and shipping routes, is having a direct impact on transportation costs across industries, including waste management.

The challenge is clear: how do you maintain service quality when your largest variable cost is skyrocketing? The answer lies in a combination of transparent billing, technical efficiency, and data-driven route management.

Why it matters:

Fuel can account for up to 30% of municipal fleet operating costs
Rising costs can trigger fuel surcharges passed to clients
Inefficient collection routes exacerbate the problem

The 2026 fuel crisis: what’s happening and why it matters

In March 2026, the average retail price of diesel in the U.S. climbed above $4.04 per gallon, an increase not seen in nearly two years. Analysts suggest prices could continue rising, possibly reaching $4.25–$4.45 per gallon depending on developments in global crude markets.

The primary driver of this spike is a continuation of conflict in the Middle East, including retaliatory strikes and disruptions to oil infrastructure that have tightened global diesel inventories and shipping flows. Supply chain bottlenecks, especially through critical chokepoints like the Strait of Hormuz, through which approximately 10% of the world’s diesel supply moves, have increased sensitivity to these geopolitical events.

This means diesel price increases aren’t isolated to consumer gas stations: they ripple throughout the industrial economy. Waste collection companies, heavily dependent on diesel for daily routes, find themselves on the front lines of cost pressure as fuel becomes more expensive.

Why fuel costs matter in waste collection

Waste collection fleets operate under unique cost structures: they involve heavy vehicles, frequent start‑stop driving, and long daily distances traveled. Unlike passenger vehicles, commercial waste trucks rarely benefit from efficient highway cruising; instead, they spend much of their time in slow, congested urban environments, where idling and stop‑and‑go traffic further increase fuel waste.

Moreover, rising diesel prices don’t just affect direct fuel purchases. Higher diesel increases costs across the supply chain, from landfill operations to processing facilities, and can cascade into higher service fees if not managed proactively.

Route Optimization: the most effective immediate lever

Before implementing any cost-saving strategy, fleets must understand how fuel is consumed across their operations. Telematics and onboard diagnostics provide real-time data on fuel consumption, idle times, mileage, and driver behavior. Such data helps identify inefficiencies and allows fleet managers to prioritize the most impactful interventions. Analyzing route patterns can highlight redundant mileage, unnecessary detours, or inefficiencies caused by urban congestion, while tracking driver habits can reveal excessive idling or aggressive driving that increases fuel consumption.

Route optimization emerges as the single most effective lever to reduce fuel expenses. Smart route planning considers traffic conditions, stop sequencing, vehicle load, and time windows, minimizing unnecessary mileage and idle time. Studies suggest fleets implementing route optimization can reduce total miles driven by 10–30%, with corresponding fuel savings of 15–20%. The financial impact scales with fleet size: even a 10% reduction in mileage for a mid-sized fleet can save tens of thousands of dollars annually.

How route optimization saves fuel:

Eliminates redundant mileage: Fewer backtracking and overlapping routes.
Reduces idle times: Avoids rush hour zones and stop‑heavy traffic segments.
Enhances scheduling: Groups stops in the most fuel‑efficient geographical order.

From static schedules to dynamic intelligence

Historically, waste management has relied on “static routing”—the practice of following the same map every Monday, regardless of whether a bin is overflowing or completely empty. In a pre-2026 economy, the cost of “collecting air” was a tolerable inefficiency. Today, with the Middle East conflict pushing diesel toward the $5 mark, collecting an empty bin is no longer just a minor waste; it is a direct hit to the company’s solvency.

To bridge the gap between rising costs and operational survival, the industry is shifting toward Demand-Responsive Collection (DRC). This is not just about better maps; it is about real-time visibility into the infrastructure itself.

Enter NANDO.Sensor: The digital hedge against volatility

The most significant breakthrough in this space is the deployment of IoT-enabled hardware like NANDO.Sensor. By installing these rugged, ultrasonic devices inside commercial and residential containers, haulers can finally stop guessing.

NANDO.Sensor works by constantly monitoring the fill-level, temperature, and orientation of the bin. This data is transmitted via cellular or LoRaWAN networks to the Nando AI cloud, where it is transformed into actionable intelligence.

The “Zero-Waste” manifest: Instead of a driver starting the day with a list of 500 stops, NANDO generates a dynamic manifest of only the 320 bins that actually require service.
Preventing the “Emergency Run”: High fuel prices make unscheduled, one-off pickups (due to overflows) incredibly expensive. NANDO uses predictive analytics to alert managers before a bin reaches capacity, allowing it to be integrated into an existing fuel-efficient route.

When we talk about waste hauler fuel efficiency, we are really talking about the “Cost Per Lift.” In a high-diesel environment, the only way to keep the Cost Per Lift stable is to increase the volume of waste collected per mile driven.

NANDO.Sensor facilitates this by maximizing “bin density” per route. By ensuring that every stop on a route results in a significant volume of waste collected, the fuel consumed by the truck’s hydraulic compaction system is balanced against a higher payload.

Furthermore, the integration of sensor data with NANDO’s route optimization engine addresses the “Idling Problem.” Garbage trucks burn approximately one gallon of diesel for every hour of idling. By using real-time data to avoid congested areas and optimizing the sequence of stops to minimize PTO (Power Take-Off) engagement, Nando AI helps fleets reclaim hours of wasted fuel every week.

The new standard for 2026 and beyond

The era of cheap, stable energy is over. Geopolitical tensions in the Middle East have proven that the waste industry can no longer afford to be “analog.” To survive the $4+ gallon reality, haulers must embrace a philosophy of Precision Waste Management.

By combining the rugged reliability of NANDO.Sensor, companies can move from a position of vulnerability to one of resilience. You cannot control the price of oil, and you cannot control global conflict but you can control every mile your trucks drive.

The choice is simple: continue to pay the “inefficiency tax” at the pump, or invest in the intelligence that makes every gallon count.

Is your fleet ready for the $4.50/gallon threshold?

Don’t wait for the next price spike. See how NANDO.Sensor can war-proof your operations today.

See How it Works

The post Waste collection route optimization with AI: how to cut costs now with NANDO appeared first on NANDO.

How Dussmann reduced canteen food waste by 60% with AI

Maria Balvis — Tue, 24 Feb 2026 14:35:55 +0000

Home

How Dussmann Reduced Canteen Food Waste by 60% with AI — Case Study

Dussmann Service cut food overproduction waste from 20% to 6–8% in a corporate canteen in Rome using NANDO.Canteen AI cameras. Here is how AI-powered food waste monitoring transformed their meal service, and what it means for any food service operation in 2026.

Key results at a glance

Metric	Result
Food overproduction waste	–60% reduction
Uneaten food rate	20% → 6–8%
Recycling accuracy improvement	+60% in 2 months
Data categories monitored	21 food categories

The Challenge: Food Waste Without Visibility

Corporate canteens face a structural problem: food is produced based on estimates, not data. When those estimates are wrong, which they frequently are, the result is systematic overproduction, food waste, and avoidable costs.

Dussmann Service, one of Europe’s leading facility management companies, faced this challenge in their corporate catering operations. Without real-time visibility into what diners were actually eating versus leaving on their trays, kitchen staff had no reliable basis for adjusting portions or menus.

The consequences were measurable: high food waste rates, elevated food costs, and an ESG gap in sustainability reporting.

The Solution: NANDO.Canteen AI Food Waste Monitoring

Dussmann implemented NANDO.Canteen in a corporate canteen in Rome. Two AI cameras were installed strategically at the tray return area, the point where uneaten food becomes waste data.

How NANDO.Canteen Works

1. 1. Automatic detection: Cameras recognize food residues across 21 categories (pasta, meat, vegetables, fruit, bread, etc.)
1. 1. Weight estimation: Advanced algorithms calculate the precise quantity of food wasted per tray
1. 1. Real-time dashboard: Data aggregated and visible immediately to kitchen staff
1. 1. Weekly reports: Actionable insights for menu and portion optimization
1. KPI monitoring: Continuous tracking of waste evolution over time

The system requires no changes to kitchen workflows or diner behavior — it operates passively, turning every returned tray into a data point.

Results: From Gut Feel to Data-Driven Decisions

Food waste reduction

Within months of deployment, overproduction waste dropped from 20% to 6–8% — a reduction of more than 60%. Kitchen staff could now identify with precision which dishes were consistently left uneaten, and adjust production quantities accordingly.

Recycling quality improvement

Alongside food waste reduction, NANDO.Canteen’s monitoring contributed to a 60% improvement in recycling accuracy within the canteen in just two months — a direct result of better visibility into waste streams.

Menu optimization

By analyzing real consumption data — what was actually eaten versus what was left — Dussmann was able to:

- - Eliminate dishes that were systematically unpopular
- - Increase frequency of highest-rated meals
- - Test new recipes with immediate feedback from consumption data
- - Adjust portion sizes to match actual diner preferences
- Personalize menus across different corporate locations

Economic and environmental impact

Cost savings

Reducing food overproduction waste by 60% translates directly into lower food costs. Less raw material is purchased, less staff time is spent on excess preparation, and less waste requires disposal.

ESG and sustainability reporting

Every kilogram of food saved with NANDO.Canteen represents less water, energy, and natural resources consumed in food production — and lower CO₂ emissions from both production and organic waste disposal. This data feeds directly into ESG sustainability reports.

As Alessio Vismara, COO of Dussmann’s Business & Industry division, noted:

the system allows analytical, accurate monitoring of perceived quality within their corporate restaurants. improving service and better satisfying client preferences.

Circular economy integration

The food waste data generated by NANDO.Canteen also opens pathways for surplus food recovery programs and circular economy certifications, increasingly relevant for corporate sustainability strategies under CSRD.

Why this model is scalable

The Dussmann implementation demonstrates that AI-powered food waste monitoring is not limited to large corporate canteens. The same approach applies to:

- - Hospital and healthcare catering — where dietary requirements make overproduction particularly costly
- - University and school canteens — high volume, variable attendance, significant waste potential
- - Airport and transport hub food service — unpredictable demand patterns where data is especially valuable
- Commercial restaurant groups — multi-location operations needing centralized waste visibility

NANDO.Canteen deploys in under 48 hours per location, requires no structural changes, and begins generating actionable data from day one.

Frequently Asked Questions

How does AI reduce food waste in corporate canteens? NANDO.Canteen uses cameras at the tray return area to identify and weigh uneaten food across 21 categories in real time. This data allows kitchen managers to adjust portions and menus based on actual consumption patterns rather than estimates.

How quickly does NANDO.Canteenshow results? Dussmann saw measurable waste reduction within the first weeks of deployment. Overproduction waste dropped from 20% to 6–8% within months, and recycling accuracy improved by 60% in the first two months.

Does NANDO.Canteen require changes to canteen layout or workflow? No. The system uses cameras installed above the tray return area. No structural changes, additional hardware at table level, or modifications to kitchen or service workflows are required.

Can food waste monitoring data be used for ESG reporting? Yes. NANDO.Canteen generates certified food waste data compatible with CSRD, GRI, and corporate ESG frameworks. Data is automatically aggregated and exportable for sustainability reports and audits.

Is NANDO.Canteen suitable for high-volume food service environments? Yes. NANDO.Canteen is designed for high-volume operations including corporate canteens, hospital catering, university cafeterias, and large-scale food service. It operates in real time without slowing service or requiring workflow changes.

Ready to Reduce Food Waste in Your Canteen?

Discover how NANDO.Canteen can help you monitor, analyze and reduce food waste with AI technology, just like Dussmann did.

Book a Demo
Other Case Studies

Related pages: NANDO.Canteen · Food Waste Industry · Case Studies

0 Shares

The post How Dussmann reduced canteen food waste by 60% with AI appeared first on NANDO.

Digital Waste Tracking: why manual scales cost you more

Maria Balvis — Mon, 23 Feb 2026 16:11:19 +0000

Digital Waste Tracking: why manual scales cost you more

Home

Manual waste weighing with a mechanical scale is still the most common method for collecting data on waste generated in commercial, industrial, and service buildings. It is a consolidated process, familiar to operators, and seemingly reliable.

The problem is that the data it produces is less reliable than it appears, and the operational cost required to obtain it is much higher than typically calculated.

In 2026, this limitation has become harder to ignore. Companies face at least three concrete pressures that require accurate, digital, and auditable waste data:

ESG reporting obligations (CSRD): Large European companies must produce transparent and verifiable data for sustainability reporting, including waste data and Scope 3 emissions.

ISO 14001 and EMAS: Companies certified under ISO 14001 or registered under EMAS must monitor and measure their environmental impacts, including waste generation, with data that can be verified by external auditors. Manual, non-calibrated measurement does not meet the reliability requirements of these standards.

Circular economy and corporate sustainability strategies: Achieving zero-waste targets or waste reduction goals requires high-granularity data on waste composition, not just total weight. Without knowing what is being discarded and where, it is impossible to define an effective strategy.

An analog scale produces a number. Today’s regulations require data.

The Traditional Process: Four Steps, Four Points of Error

The standard measurement flow using a mechanical scale consists of four stages:

Manual weighing of the bag on the scale
Manual recording of the weight on paper or in a digital register
Data entry into Excel
Sharing the report with the manager

Each step introduces a margin of error. None of the four produces disaggregated data by waste type. And the entire process requires operational time that is rarely accounted for in the real cost of waste management.

The real issue is not weighing — it is tracking. The manual process creates a lack of real-time alerts, no visibility into contamination, and no ability to forecast future volumes. The data exists only at the moment it is written down and then disappears into Excel until the monthly report.

Most Common Types of Mechanical Scales

Several types are available on the market, with different features and tolerances:

Spring scale (±5–10% tolerance)
Bench scale for light waste
Industrial waste scale (capacity up to 500 kg)
Cart-integrated collection scale
Dedicated scale for canteens and retail (GDO)

All share the same structural limitation: they produce a total weight, without information about waste composition or quality.

The Accuracy Problem: Why 78% Is the Real Figure

A new, calibrated mechanical scale has a standard tolerance of ±0.1 kg (OIML R111 standard). On a 4 kg bag — the average size in commercial or industrial contexts — this corresponds to an initial 3% loss in accuracy, which may seem acceptable.

The problem emerges over time and in real operating conditions:

ISO 9001 calibration. The ISO 9001 standard requires certified annual calibration. Without it, tolerance can increase to ±0.5 kg after only six months of continuous use, reducing accuracy by approximately 9%.

Human error. Incorrect bag placement, double counting, misreading, distraction — these factors cause variations of up to ±10% in real operating conditions.

Adding these factors together:

Accuracy Reduction Factor	Mechanical Scale	NANDO.App
Standard tolerance	-3%	-5%
Missing ISO 9001 calibration	-9%	ISO 9001 certified ✓
Human errors	-10%	0
Real final accuracy	~78%	~95%

A non-calibrated scale used under real conditions averages 78% accuracy. NANDO.App, by eliminating human error and ensuring ISO 9001 Bureau Veritas certification, reaches 95%.

The difference is concrete: on a site producing 100 kg of waste per day, 78% accuracy means 22 kg of incorrect data every day — approximately 8,000 kg per year of measurements on which operational, reporting, and sustainability decisions are based.

How NANDO.App Works: A Photo Instead of a Scale

The operator installs NANDO.App on the smartphone they already use. The process is reduced to two steps:

Take a photo of the bag contents before replacing it
Automatic AI-generated report on the centralized dashboard

Main system features:

Object recognition: AI identifies what is thrown away, category by category
Weight-volume association: advanced algorithms correlate visible volume with estimated weight
Contaminant detection: the system identifies misplaced materials that reduce recycling quality and increase disposal costs
Real-time dashboard: data available per bin, floor, building, or multi-site network
Certified ESG reporting: auditable data compatible with CSRD, RENTRI, MUD, and major international frameworks

No scale. No paper logs. No connection to the site’s private network: the app runs autonomously on 4G.

The Hidden Cost of Manual Weighing

The comparison is not only about accuracy. The operational cost of manual weighing is systematically underestimated.

Data collected from real NANDO installations with 300 bins shows:

Measurement Activity	Without NANDO.App	With NANDO.App
Time per bag	1 minute	6 seconds
Total daily time	5 hours/day	0.5 hours/day
Labor cost (30$/h)	$54,000/year	$5,400/year
Net savings	—	$48,600/year (85%)

The 85% savings in labor cost dedicated to waste measurement emerges from comparing one minute per bag with a mechanical scale versus six seconds with NANDO.App — translating into tens of thousands of dollars per year for sites with medium waste volumes.

Accuracy That Improves, Not Degrades

A structural difference compared to mechanical scales: NANDO.App’s accuracy does not degrade over time — it improves.

Each photo feeds a dynamic database updated in real time through generative AI. The more the system is used in a specific context, the more precise it becomes for that environment.

Internal data shows that in the first 12 weeks accuracy increases progressively from an initial 60% to 92%, where it then stabilizes. A non-calibrated mechanical scale follows the opposite trajectory.

The Data a Scale Cannot Provide: Waste Composition

Even under optimal conditions — perfectly calibrated scale, no operator error — a fundamental limitation remains: total weight does not reveal waste composition.

A weighed bag returns a number. It does not say whether the waste is paper, plastic, organic, or mixed. It does not identify contaminants. It does not allow calculation of recycling quality.

By classifying bag contents through AI, NANDO.App produces disaggregated data by material type, enabling analyses impossible with mechanical weighing: segregation quality, contaminant identification, collection optimization by stream, ESG and CSRD Scope 3 category reporting.

Which System Should You Choose?

The answer depends on the operational context and regulatory obligations.

Manual weighing with a mechanical scale remains sufficient for very small waste volumes, no ESG reporting requirements, and no need for category-level data. In all other cases, the structural limitations of the system — degrading accuracy, non-auditable data, high operational cost — become a concrete problem.

NANDO.App is the appropriate choice for organizations managing multiple sites, subject to CSRD or RENTRI obligations, aiming to optimize recycling with real data, or needing to reduce operational time dedicated to measurement. Digitalizing waste measurement is no longer optional for those who must transform waste weight into actionable business intelligence.

Replace the scale with a photo.

Discover how NANDO.App can reduce manual measurement costs and provide finally reliable waste data.

Book a free demo

Frequently Asked Questions

For spot measurements or an initial baseline, manual weighing still makes sense. Its limitations emerge when used as the sole continuous monitoring system: degrading accuracy, aggregated non-auditable data, and high operational cost compared to informational value.

No. The workflow is reduced to taking a photo of the bag before replacing it. No manual registration, no technical expertise, no access to site IT systems is required.

Initial accuracy ranges between 60–70% and increases over the first 8–12 weeks, reaching 92–95% as the AI model learns the specific operational context. After 12 weeks, accuracy stabilizes.

Yes. The platform produces ISO 9001 Bureau Veritas certified, continuous, and auditable data compatible with CSRD Scope 3 reporting and major international sustainability frameworks (BREEAM, LEED, EMAS, ISO 14001).

The post Digital Waste Tracking: why manual scales cost you more appeared first on NANDO.

NANDO Raises €3.3 Million to Innovate Waste Management

Maria Balvis — Mon, 16 Feb 2026 16:40:33 +0000

NANDO Raises €3.3 Million to Accelerate Innovation in Waste Management

Funding round led by MAIA Ventures and CDP Venture Capital to support international expansion and AI development

Key Highlights of the Funding

€3.3M Investment

Capital raised for international expansion and AI technology development

17 Countries Served

Active clients across Japan, USA, UAE, Europe, and UK

40,000+ Images/Day

Waste data processed for data-driven analytics

80+ Clients

Collaborations with leading international companies

Data-Driven Innovation in Waste Management

NANDO has successfully closed a €3.3 million funding round, marking a strategic milestone to consolidate its technological leadership in the waste management sector. The investment will accelerate international expansion and develop new AI-powered features.

The round was led by MAIA Ventures and CDP Venture Capital SGR, with participation from existing investors including Club degli Investitori, La4G, and EMBA Capital Partners. This confirms the robustness of NANDO’s business model and the potential of AI solutions to transform waste management processes.

Relevant Data

NANDO currently processes over 40,000 images daily, serves approximately 80 clients in 17 countries, supporting major companies such as ISS, Narita International Airport, Samsic Facility Italia, A2A, IREN, and IVECO Group.

Expansion and AI Technology

The raised funds will strengthen AI technologies, develop new product lines for different types of waste, and expand the NANDO team. A key strategic focus is on food waste management in collective catering, where the recently launched solution is already receiving strong market interest.

NANDO’s platform leverages AI, Machine Learning, and Computer Vision to transform waste into actionable, granular data, enabling companies to digitalize, automate, and optimize their management processes with measurable reductions in time, cost, and environmental impact.

Vision for the Future

“The capital raised will allow us to further develop our technology, solidifying NANDO as a single access point for waste data.”
— Riccardo Leonardi, CEO & Co-founder

NANDO’s mission remains clear: putting data at the center of waste reduction strategies, operational optimization, and decision support. Less waste, more efficiency, greater sustainability.

Want to Optimize Waste Management in Your Company?

Join over 80 leading companies that have already chosen NANDO to digitalize and optimize their waste management processes.

Book a Free Demo

The post NANDO Raises €3.3 Million to Innovate Waste Management appeared first on NANDO.