The impact of generative AI on the waste management industry

Dr Tracey Leghorn
by Dr Tracey Leghorn, Chief Business Services Officer | SUEZ recycling and recovery UK
Just about everyone agrees artificial intelligence (AI) is, or will be, a transformative technology. How generative AI and other digital solutions will re-shape our working lives, and the waste management industry, is still unclear. But we’re already seeing change, and it’s accelerating.

Accurate sorting is pivotal for maximising production of high-quality secondary raw materials. Manual sorting has proven challenging, especially in plastics recycling. At SUEZ, we are continuously enhancing our waste-sorting techniques by leveraging cutting-edge technologies such as optical sorters with high-definition scanners, mechanised biological treatment, trommels, ballistic screens – and yes, sorting robots.

The integration of such advanced technologies has enabled a significant enhancement in the productivity and performance of our sorting centres. In our latest facilities, we have achieved a remarkable recovery rate of up to 90%, generating top-grade secondary raw materials that meet or exceed the stringent requirements of our industrial clients.

In future, with real-time analysis and near-total insight into the purity rates of specific bales, AI data can more accurately inform pricing and encourage the use of quality recyclates over virgin resources. We can also improve transparency and thus, customer communication with our clients, which is central to our business success.

Other AI applications

There are important applications too in our energy-from-waste (EfW) and alternative fuel manufacturing operations. Automated monitoring and predictive maintenance bolster operational efficiency by minimising unplanned shutdowns. Continuous assessment and analysis of equipment health and productivity afford valuable insights for timing maintenance checks, part replacements, and recalibration – pre-empting equipment failures and avoiding the disruption and cost of emergency repairs.

Our latest energy-from-waste facilities employ AI-computer vision as an extra set of eyes to detect non-conforming waste before it causes blockages and shutdowns. In its first four months, the system detected 11 non-compliant items missed by the operator that could have caused systems shutdowns.

Since Brexit, our industry has experienced a shortage of machine operators and manual staff. The CIWM has forecast approximately 10,000 job vacancies for roles in process, plant, and machine operation before 2029. The automation of repetitive tasks can alleviate this shortfall. It will also improve the occupational health of the people operating these AI-enhanced machines.

Informing policy-makers

What might this new intelligence mean at national level? Manual sampling, which typically covers just 1% of the waste stream, informs government policies as well as operational decision-making. With optical scanners and AI software analysis, we may achieve over 99% visibility. Knowledge of the totality of the waste stream will be crucial for the transition to a circular economy.

One example is the tracking of commodities in residual waste to prevent unnecessary diversion to EfW and landfill. In 2023,a GreyParrot study that analysed residue streams across Europe, the US, and Asia discovered a staggering 37% was recyclable paper and cardboard, and recyclable plastics made up another 26%. In-depth insights into these lost resources should lead to strategies (also AI-powered) to recover them in real time in the future, whether for conversion into industrial fuels or other valuable uses.

Mining our data

Despite being a sector rich in data, the waste management industry could improve both data collection methods and the analysis of the data we do gather. AI analysis offers a high level of accuracy and efficiency in processing large volumes of data, enabling valuable insights into patterns, trends, and correlations that may not be immediately apparent to human analysts. Information captured and analysed with the help of AI can plug our sector into the entire value chain, from product development to sales, so end-of-life product information can inform decisions about eco-design, re-use, repair, and recycling initiatives.

However, as with AI applications in other fields – from CCTV to CV screening – there is potential for bias in the algorithms, particularly if the training data doesn’t prepare the system for accurate recognition. In the recycling and resource industry, this will mean mis-categorising materials, potentially skewing results. The risk of misidentifying unusual items adds to the need for human oversight and some manual sorting.

AI will put a higher premium on some human skills. Trained staff are needed to interpret the results of data analysis and counteract any potential bias (or AI hallucinations). We also need to design and implement comprehensive privacy and security measures. The future will also necessitate greater collaboration between our industry specialists and AI experts (some of whom may be in-house). The influx of AI and machine learning will bring new types of workers into the industry, with less expertise in waste management but greater data and IT skills.

HR professionals have a key role to play in blending these capabilities as well as upskilling and re-skilling our people for changing roles, while also addressing emerging concerns about this brave new world of smarter working. The use of AI must be explainable and subject to rigorous human oversight. Robust AI governance is required company-wide and it must be ethical, risk-focused, and adaptable as the technology rapidly evolves.

A people-centred approach

This transformation should be consistent with our human-centric approach and ‘growth mindset’ as we create fulfilling jobs, encourage our people to innovate, embrace digital challenge, and pursue their professional development.

We’re still in the hype phase of AI. The robots were coming to take all our manual jobs. Will AI now render our data collection, analysis and decision-making skills redundant? Or, to be more of a techno-utopian, will these convergent technologies solve our most vexed problems, like low recycling rates in tower blocks? Imagine a self-contained waste processing and recycling unit in the basement of every building that automatically sorts, cleans and weighs waste materials as they arrive via a chute from the flats above, leaving them baled and ready for shipment to reprocessors?

Maybe not. But if we (carefully) embrace the transformative potential of AI, it could be a powerful ally of an innovative industry with agile and skilled people empowered to make the circular economy and sustainability a reality.