To view the original article, please click here.
JD Ambati and Don Gambelin
Without actionable intelligence provided by Google, Walmart has no idea how well their products are selling on their website. Likewise, Walmart has its own data that guides Google about making adjustments to advertising spend and strategies based on sales. What does that have to do with Materials Recovery Facilities (MRF) operators? The answer is a lot.
If Walmart is at the epicenter of retail sales and consumption, MRF operators are at the epicenter of making our recycling programs successful worldwide. Every recycled object flows down their conveyor belts. For any ‘green, environment, social and governance (ESG), extended producer responsibility (EPR), climate change goal’ initiative to be successful, we need the MRF operators to be successful first. The question is, have we empowered MRF operators – owners and staff – with technologies to capture every single recyclable object?
The answer is no. To date, MRF operators have lacked modern tools and analytics that powerful companies like Walmart, P&G, Ford, Google, Facebook and Apple use to run their businesses. On top of that, there has been a glaring gap between how MRFs are run, what the governments want, what manufacturers seek with recovery and procurement, and what consumers expect from their actions.
The USEPA estimates that we are losing $3B worth of aluminum to the landfill every year. But how are we losing it? Where? When? Why? How many cans at each plant? And how many on each conveyor belt? To close the gap in achieving our green, ESG, EPR, climate change goals we must connect the dots between packaging manufacturers, brands, MRF operators, consumers and governments.
Running an optimized recycling system can be achieved when we employ innovative digital tools that support extended producer responsibility with transparency and are driven by metrics that measure recovery to the single object level. If Google can measure return on ad-spending down to the penny, then we should be able to tell what happened to every container that got manufactured, consumed and put into the recycling ecosystem. And that is the holy grail – being able to measure performance down to the object level.
Knowledge down to the object level has been one of our core tenants. We believe that the artificial intelligence solutions being used by the world’s most successful companies can and should be applied to recycling. Experience tells us that AI solutions can be built from the ground up for the recycling ecosystem by spending time with the MRF operators and by listening to MRF operators’ pain points from all angles that present core challenges in optimizing their operations. We believe stand-alone AI computer vision systems can be deployed across the industry.
Minute-by-minute tens of thousands of objects on MRF conveyor belts are being counted and identified to empower MRF operators with real-time knowledge of operations, revealing both the good news and the bad. This type of data is auditable and transparent and can be used to facilitate transactions for EPR and ESG.
And, importantly, data should be used to drive decisions such as automating activities with machines and systems (robotics and more) and enabling MRF staff in taking proactive steps to fix equipment issues before they break down, grinding MRF operations to a halt. The data can help MRFs stem the losses of tens of thousands of dollars’ worth of recoverable recyclables to the landfill every day – resulting in revenue leakage.
We believe software should act as an early warning system — or your wingman who never sleeps. Customers should be able to start with their own data to build from. The key is customizing automation tools to plant needs, not force-fitting a one-size fits all approach to AI powered equipment.
ReThink Waste in San Carlos, CA, invested in substantial equipment upgrades and optimized their recovery over a few months. Once we installed our AI vision systems, our data provided invaluable feedback on the effectiveness of their investment, how manual human resources should be redeployed based on the equipment’s performance, and additional insights for the next phase of upgrades. Our AI documented the increase in MRF production, showing a 300% increase in material flow.
Our AI automatically quantified that the customer was able to reduce the leakage of revenue from their new equipment – i.e equipment removed valuable bottle bill materials from the fiber stream. On a daily basis, this customer significantly reduced lost revenue (leakage costs) through increased recovery documented. In the same plant; our AI software on a last chance line identified a net new revenue opportunity for the MRF by identifying a large number of containers leaking through the MRF equipment. Using this information, our AI platform calculated the value of leaked containers and resources needed to recover them and informed operations which containers should be prioritized for recovery based on value and carbon footprint.
Our AI technology is deployed on a last chance line in a high volume MRF in Northern California. In this MRF, our software is used for analyzing and monitoring hundreds of data points on the conveyor line every second. From these data points our software provided actionable insights about the need for maintenance of the upstream equipment as well as the increasing amount of revenue leakage. As per this customer, they would have taken weeks or even months to identify the issues brought to light by our software due to manual processes and reactionary behavior involved by maintenance staff as the issues are not obvious to the naked eye.
We need to empower and support our unsung heroes – MRF operators – at the epicenter of recycling to achieve our ‘green, recycling, ESG, EPR, climate change goals’ by measuring their daily incoming volumes down to the object level. Our ultimate objective for the collective good: making sure no recyclable is left behind.
