Recently I downloaded a free app called IFTTT (if this, then that) in order to sync some of the apps I use like triggering the thermostat to turn down when another app recognizes that the last person has left the house.
This go me thinking about the concept of IFTTT and it’s role in waste management, specifically sorting. By now, if you didn’t already know, your likely confronting issues around contamination in your waste stream, especially for multi-family real estate operators.
Contamination is the Kryptonite of recycling, an issue that was highlighted by China when they announced they would drastically reduce the amount of recycling they would accept that included contamination. Contamination effects recycling in several negative ways:
When a load of recycling includes too high of a percentage of contaminants the entire load is sent to the landfill as the extra costs to separate the contaminants exceeds the value of the recyclable resources in the load.
Particular contaminant items (i.e. PLASTIC BAGS) can wrap around shafts and axles of sorting machinery not only shutting down the entire line, costing time and money, but also potentially endangering the sorters who have to remove them.
The presence of hazardous waste when included in a recycling load can not only contaminate other items but expose recycling workers to disease and other physical harm.
The value of the resource itself can be reduced or eliminated, driving down the value of recycling and ultimately harming the entire recycling program. This may also result in increased costs for the recycling service.
The relationship between the recycling facility and sorter can also be damaged, eventually the facility grows weary of continued receipt of contaminated loads and may start to refuse to receive loads from a particular hauler - really no different then what China did with their China Sword Policy. This results in that haulers loads ending up in a landfill.
So what does this have to do with IFTTT? Through the power of machine learning and platforms such as Python with a fastai library, convolutional neural networks are being developed that can identify images and place them into the correct category. In fact, in March of 2019, Collin Ching outlined just such a project he put together, “just for kicks” - you can read about it here: How to build an image classifier for waste sorting.
To this point, most of this technology has been deployed at the MRF level, but increasingly haulers are deploying similar technology at the site of collection to help identify contamination before it impacts the entire load. WM SmartTruck offers an example of the combination of cameras, trucks and technology being used to look for contamination as well as other conditions including overages, damaged containers, graffiti, recording service levels and more.
The next logical step is taking this technology to the computer that you hold in your palm, your phone. Through the development of a app that could be used by onsite sorters, door to door service workers or even a consumer facing app. The ability to recognize what item goes where, when it rests in the form of a downloadable app holds tremendous promise.
Currently we tend to “train residents” through signage and perhaps a website that they have to go to and search to figure out how to handle a particular item. While these strategies certainly have a place, neither makes recycling particular easy. We reinforce that training with curbside feedback in the form of tags or fines and potentially an annual mailer. Let’s face it, all of the measures rely a lot on the resident remembering or taking the time to look for how to do something.
Yet, if a consumer wants to know if a food fits into their daily calorie allowance, they simply take out their phone - scan the bar code and “voila”, the nutrition content is not only displayed but can be conveniently placed into a daily meal log for you.
The challenge is before you waste industry professionals - engage the resident and sorter, provide them a tool that recognizes and advises where an item can go. Collin Chang literally developed a prototype on a whim, surely with some planning and purpose as an industry we can do the same. Who will emerge as a technology leader?
Are any of you currently engaging your residents on how to recycle? If so, what tools are you using and how are you deploying them? I look forward to continuing the conversation either here or directly.
Stay well!