Garbage Collection, Without the Noise or the Smell

The VakTrak doesn't look like dumpsters on wheels, but it's one of the models the MTA uses for track cleaning: https://www.youtube.com/watch?v=GBSfb84z4QI

Edit: I think you saw an R127 hauling garbage: https://www.youtube.com/watch?v=a6VnQzShTZg. I've never seen this before personally, but I guess the MTA does haul garbage ;)

http://www.nytimes.com/2011/10/25/nyregion/mta-tries-to-redu...

The truth is that the garbage train only collects garbage at the stations. Not the entire city. Think about it, when it comes time to remove the trash in the station dumpsters, why run a truck over ground and lug trash up stairs when you can just run a flat-bed down the tracks.

Other unusual subways include the rare "track inspection vehicle" and the now discontinued "money train".

I of course agree with your sentiment here. It scares me how general population has everything exactly backwards when it comes to aggregate effects. Fighting centralized sorting is one example. Fretting over unplugged chargers is another (savings here are minuscule; most people would save orders of magnitude more of energy if they stopped driving their cars so much and used public transport every now and then). It's as if people cared only about gestures and feeling good about themselves, and not about actual effects.

(I've had a previous landlord that was very into recycling and he would remind us to take the caps off of bottles before putting bottles in recycling. Either he was wrong, or most people I see recycle bottles wrong. I also went to a very liberal high school that was super into recycling, and had large central sorting bins, but half the large recycling pickups were rejected for being too impure.)

Here in CA, South Tahoe Refuse (STR) uses a modified process that I think solves a lot of that problem. They give their customers durable semi-transparent blue bags to fill with recyclable materials. This is somewhat ceremonial but does certainly aid in the sorting efficiency. The waste generator then places those bags in with the regular garbage and one truck collects it all. At the STR operated MRF, they mechanically separate the blue bags and break them open. The resulting recycling is then hand sorted. From what I’ve seen at their MRF, the diverted material is of much higher quality than materials I’ve seen at “regular” dirty MRFs.

I’d personally love to see waste collection return to single container/vehicle or some less impactful method. The current regulations in CA essentially require most locations to run 3 vehicles that collect from 3 different containers at each residence/business since you have to hit a certain level of diversion (keeping materials out of the landfill) that is hard to attain at a dirty MRF. Add to that the CA health code requirements for weekly collection and you can’t just “alternate” collection. The additional fuel consumption, road wear, and other resource demands that result from this multiplication of equipment & travel would be super interesting to research and document. Figuring out a good solution that would allow you to centralize sorting and hit high diversion percentages would result in less pollution/resource demands and potentially lower rates for the waste generators.

Also, now everyone else is experiencing my inverse experience to every “garbage collection” article on HN. I’d love to see more innovation, thought, and exploration around the problem of physically collecting garbage and I love when it’s discussed here on HN.

Ultimately, the stuff being recycled does have to be separated in order to be made useful again. Doing this centrally lets you automate some of that, but you'd be surprised/depressed at how much of the separating is done by hand. There's no magic separator robot at the end of the line.

Also, the quality of the result goes down. Bits of broken glass get mixed into the paper. This makes the end product less valuable, which sucks money out of the entire pipeline. If it's cheaper to single-stream recycle, but the recycled product gets sold for half the price, you haven't really gained anything.

Some more details: http://www.theatlantic.com/technology/archive/2014/09/single...

In constrast, it takes a lot less effort to separate the bottles from the paper on my kitchen counter where they are already sitting there clean and separated. Single-stream recycling to me seems like letting users add a ton of entropy to the system which then has to be laboriously removed later on. Why not just leave the chaos out to begin with?

What I do think is a real problem with decentralizing and letting everyone participate a bit in the recycling process is that there's no feedback loop. I spend an embarrassing amount of mental anguish on trying to decide if some piece of paper can be recycled or composted. I don't know if I'm doing it right, or how I'm doing compared to my neighbors.

I understand the information tracking would be insanely complex, but I would love some kind of feedback loop so I could do things better. Even if it was aggregated among my whole neighborhood, maybe a little friendly competition between other parts of town to get the best recycling score would help get people to participate.

Here's a fair proposal to do the opposite: At collection points throughout a city's inner suburbs have bins equipped with cheap image sensors scan a QR code on the packaging - each code containing information to identify the type of packaging as well as provide a unique code for each item so it can't be scanned multiple times. Depending on the type of packaging, a certain lid on the bin (among several) will open so that sorting is automatic, and a refund for the packaging can be either donated to a chosen charity or aggregated until a certain dollar value is accrued from future recycling deposits and paid in bulk on the spot.

Biggest limitations: getting a country to alter its packaging requirements to include an advanced bar-code system and a schema that is internationally recognized (ISO certified). The hardware side of things would be trivial, and when concentrated in the inner city you have fewer bins capturing a greater number of people's waste.

Maybe I'm overthinking it - with machine learning the bin could determine the type of bottle/can/container, open the appropriate lid, and watch as the item is surely and safely deposited. A mid-tier smart phone could do this - use similar parts including modem to connect to a server to log a credit for the user/donate to charity. Perhaps an international packaging code schema isn't required.

I personally don't mind my local four bin system (food/garden-compost/mixed-recycling/landfill) as that seems logical to me and takes basically no thought or effort and 3 of the four bins are the same size and get picked up by the same trucks on different days so they've put some thought into the whole system efficiency.

So there's some seperation there (food versus plastic bottle) but I also get to put different types of plastic and glass bottles in the same bin.

* Many checkout clerks are not familiar with properly keying in a package tare weight. This is a problem when you bring your own packaging when purchasing bulk. This is compounded when you already have some residual amount in the packaging, and want to convey the tare of the combined weight. The result is you increase the complexity of your at-home handling of packages (you have to set aside temporary space and containers for the residual amounts so your packages are empty at the store), and there is significant time at the check out for clerks looking up/asking others how to set a tare, and/or you go to customer service to correct improperly keyed-in tare values (or no tare values at all), even after politely asking the check out clerks if they would prefer you go to another line/clerk to handle the tare data. This even happens to me at supposedly highly-trained vendors like Whole Foods Market.

* Manufacturers are simply not equipped to handle consumers who want to purchase one single unit of a large quantity of product, over and over, year after year. For example, a 5 gal carboy of Dr. Bronner's soap straight from the manufacturer costs about the same as the equivalent quantity 5, 1 gal jugs from the retail store, after accounting for shipping and handling and sales taxes, respectively. That's fine, except those carboys are not cheap, and there simply isn't an option to purchase refills that ship in heavy duty collapsible plastic bags (shipped in heavy cardboard), not to speak of sending those refills back to the manufacturer. I'd love to get refills like that for my staple items, in highly-generic and reusable containers, instead of always sending containers to the recycling plant. I'm constantly amazed how prolifically we use disposable plastic packaging, when in those applications where it spends most of its lifecycle away from UV exposure, it is an extraordinarily durable material.

From a cultural perspective, I completely understand: I'm a very remote outlier consumer, and there are likely only a few thousand world wide who share my outlook and are are willing to go to some lengths to reduce how much I feed into the recycling and waste streams. My philosophy is that waste is a material expression of our lack of knowledge of how to usefully transform the embodied energy of the waste into something else we want to use, and the less waste and even recycling activity we perform, the more overall efficient we are.

I wouldn't think too much of it. It's reasonable for the likelihood of 'garbage collection' referring to memory management to be higher than collecting trash on HN. Conversely, if you saw it on CNN and thought memory management, then you might have a problem.