Probably Approximately Correct: Why AI Really Matters in Waste & Recycling (Beyond the Shiny Stuff) 

Let’s get straight to the point: everyone in waste and recycling seems obsessed with the visible side of AI – robotic arms and smart cameras sorting materials, image recognition for health & safety, or tracking trucks on their daily rounds. That’s the headline – grabbing stuff. But here’s the thing: hardly anyone’s talking about where AI could quietly deliver the biggest impact – right in the back-office, embedded in those unglamorous but utterly essential ERP and transactional processes. 

The Quiet Revolution Nobody’s Watching 

Our industry loves a spectacle – if it blinks and beeps, someone’s writing an article about it. But what about the parts of the business that don’t get a glossy case study? The billing runs, the weighbridge data, the customer queries, the endless compliance paperwork. This is where most of our time (and money) disappears, but it’s also where AI and analytics can quietly transform the way we operate. 

ERP + AI: Probably Approximately Correct – but That’s Enough! 

Let’s be honest, no ERP data is perfect. Scale, complexity, and manual touchpoints introduce noise and human-errors. But AI isn’t about perfection. It’s about being “probably approximately correct” – spotting patterns, flagging anomalies, and automating the boring stuff so your teams can focus on real value-add activities. Imagine AI agents reconciling supplier invoices, predicting missed collections before the customer calls, or highlighting dodgy data before it triggers a compliance headache. Not as glamorous as a bright yellow AI sorting robot maybe, but let’s face it, a lot more likely to boost your bottom line! 

Why the Focus on Imagery? (And Why That’s Not Enough) 

I get it, image – based AI is tangible. You can point a camera at a conveyor belt and watch magic happen. But if we only focus on what we can see, we’re missing the bigger opportunity: using AI to wrangle the invisible transactions, exceptions, workflows, and decisions that happen behind the scenes. The sector runs on thin margins and tight regulation. Any inefficiency in your ERP processes isn’t just a nuisance—it’s a risk. AI can help you spot issues before they escalate, and automate routine tasks that eat into your day. 

When should we do it? 

One thing is definite, you will do this, it is just a matter of timing. I suppose the only exception to that bold statement is a business that descends into a slow but unrecoverable decline – think Nokia, Blockbuster, Kodak etc.  

If you’re wondering whether it’s worth the investment, here’s the reality: studies show every pound/euro spent on AI agents in back-office processes can return anywhere from three to eight pounds/euros in operational savings and efficiency. In our sector, that’s the difference between surviving and thriving. And it could fund that shiny yellow robotic arm you wanted for Christmas. 

So, here’s my challenge: Let’s start a conversation about AI where it matters most—deep in the back-office. The tech is ready. The data is (probably approximately) good enough. The only thing missing is the will to look beyond the obvious. This series of blogs will take us on a journey into the detail behind these concepts. 

Are you ready to rethink what “AI in waste & recycling” really means? Let’s move past the shiny stuff and get real about the places AI can quietly transform your business.