Hacker News new | past | comments | ask | show | jobs | submit login
Galaxy Upcycling: How Samsung Ruined Their Best Idea in Years (ifixit.com)
84 points by avel on July 19, 2021 | hide | past | favorite | 31 comments



Incredible that Samsung never really launched this program. A few highlights:

> “There is another way to create even more value” than recycling, Samsung said in a video at the time. “It’s called upcycling.” With code and creativity, upcycling could turn a Galaxy S5 into a smart fish tank monitor, a controller for all your smart home devices, a weather station, a nanny cam, or lots more. Upcycling not only kept your old phone from being shredded or stuck in junk-drawer purgatory, it could keep you from buying more single-purpose devices. It was a smart way to reduce our collective upgrade guilt.

> The original Upcycling announcement had huge potential. The purpose was twofold: unlock phones’ bootloaders—which would have incidentally assisted other reuse projects like LineageOS—and foster an open source marketplace of applications for makers. You could run any operating system you wanted.

> But sometimes well-intentioned projects get muzzled inside giant companies. But that version of Galaxy Upcycling went nowhere. These days, Samsung is beta-testing an “expansion” of “Upcycling at Home,” despite Upcycling never actually shipping.

> Friends inside the company told us that leadership wasn’t excited about a project that didn’t have a clear product tie-in or revenue plan.

> The world needs fun, exciting, and money-saving ways to reuse older phones, not a second-rate tie-in to yet another branded internet-of-things ecosystem.


It's a very sad story.

> Samsung, a company without much of a public environmental message, was tossing around big ideas born at a grassroots level. This was something new. We were jazzed(...)

> Friends inside the company told us that leadership wasn’t excited about a project that didn’t have a clear product tie-in or revenue plan.

My guess: some team had a bright idea and managed to secure early associates and evangelizers before someone higher up started asking how the project actually synergizes with other company priorities. Since it obviously didn't, because it was something pro-consumer for a change, it got gutted, and the name reused to push some "value-add" IoT crapfest.


> a weather station

Galaxy S4 has a barometer, thermometer, and hygrometer. Such promise! And sad that more recent phones have ditched the latter two sensors, at least the barometer is standard now. Hopefully UV arrives soon as a sensor and air quality as well, and then soon you've got a nearly complete real weather station node.


If my memory serves well, those sensors were hilariously inaccurate. The same way generally most generic smartphone/SoC sensors are (except maybe gyro). Even as essential as a compass on recent smartphones requires constant hand waving to improve accuracy. Or try to get location in a city in airplane mode.

With dedicated sensors and an rpi, much better purpose-built stations could be built for dollars, and that may be true for mostly any other phone recycling purpose I heard lately - apart from utilising them as networked cameras, due to the good onboard processing/lens that would be harder to achieve with off the shelve parts.


> Friends inside the company told us that leadership wasn’t excited about a project that didn’t have a clear product tie-in or revenue plan.

The fix for this is probably to charge manufacturers for ewaste, so they have a financial incentive to keep devices useful.


Most plans to charge manufacturers for e-waste seem to be at the point of original sale (where money is already changing hands). In such a scenario, Samsung probably doesn't get any credit for "making the same amount of e-waste in the long run, just slightly later after the sale for a subset of the devices sold".

The way to produce less e-waste is to sell fewer units. That comes with an obvious problem for Samsung.


Switch the plan to a core charge model: Require Samsung to collect old phones for a given cost. The more of them they can keep in use, the less it costs them. We know the core charge model works: Like 98% of lead batteries get recycled.


Upvoted for an elegant solution (at least for consumers upgrading phones). I do wonder if the lead acid core charge works better than the bottle deposit because the value is more, because it's not that much extra hassle, or some other factor.

I also am not sure (and never thought about it until just now) as to where the forfeited core charge money "goes" for consumers who buy a new phone (paying the core deposit) and then don't return a core phone within the required amount of time or if they buy a brand A phone and return an old brand B phone.

But I suspect that the overall model could work pretty well with some of the details carefully thought out. (AFAIK, in the lead-acid battery case, I can't just go to a parts store and force them to buy my core for $20, but if I am buying a new battery, they are forced to take my old core instead of charging me the $20 core charge [or whatever it is nowadays].)


I feel like you could address that by whatever retailer collected the phone being both required to collect it, but then also required to send the bill to the manufacturer.

Aka, you buy an Apple phone at your carrier, you turn in your Samsung phone, and Samsung gets sent the bill for the phone's collection and recycling. This would work even if someone just wanted to drop an old phone off at the store: Samsung's still responsible for footing the bill, so stores have no reason not to collect.


There’s an initial condition problem though. It’s unreasonable IMO to retroactively create a new obligation of $25 times every phone Samsung ever made. It seems like you’d have to assess the core charge on new sales as of some date (especially if your stated intention is to encourage the design of longer-lasting phones).

If you agree to that premise, now the store has to have a way to figure out whether a given phone has a core charge refundable from the manufacturer, whether the phone is genuine, etc in order to not have the core credit they paid out rejected by the manufacturer.


That's true. Laws about ex post facto are pesky, especially when there's a dire need to address a massive oversight of harm. I almost wonder if there'd be value in government subsidizing the return of old phones to ensure they're disposed of properly, and maintain the simplicity of a return system, but I don't know if the cost of that program would be on the scale of "government budget rounding error" or "defense spending on the F-35".


Man I tried so hard to do that with my galaxy S3 which had an unlocked bootloader (I don't think it even needed an exploit.)

Despite it's popularity though the ROM quality was extremely poor later in its life. I actually built my own busybox based OS for it from the GPL kernel source on Samsung's website but couldn't get most of the peripherals to work without all the Android stuff.

I've totally given up on Android. Thanks to the mobile Linux "movement" (or whatever) there are multiple better options now.


It's mainly the OS which causes smartphones and tablets to end up on the landfill.

If it would be possible to install current, plain Linux on a device which can make use of a USB-Hub, in order to add a sound card, video camera, USB-to-serial or whatever it is, then these devices would get a second life.

Imagine a smartphone with a plain Linux on it and an extrenal USB-HDD and USB-to-Ethernet-Adapter attached to it to now serve as a DLNA server, or a simple NAS for a backup of the family's photo collections.

It's theoretically possible, but for some reason removing that Android and fully replacing it with a Linux/GNU system is close to impossible.


Honest question - wouldn't rooted android phones serve the main use cases of what you're proposing?

To me, it seems the landfilling is mainly because people aren't technical and/or don't have any need for a tiny linux computer.


Raspberry Pi has proven that people can find uses for tiny low-power computers if it's hacker friendly


The issue is that you've still got Android's Java VM running alongside your lean Go/Rust/Python/Ruby servers with close to no use at all.

To me it's an entirely different experience to code a server in Python and leave it there running as a background process than to code an Android app which serves the same purpose, the latter is unnecessarily hard, specially keeping that process alive over weeks.

I love my Raspberry Pis because I can do whatever I want with them, I've got full access to the system and no junk is preinstalled.

If a phone is rooted, you can get the same kind of access, but the entire experience is so much different, even if you use root to run processes directly on the shell.

LineageOS might be excellent for devices which could still serve as phones, but nearly all of my old devices are just too old to offer an experience similar to modern phones, but they could well serve other purposes.

At that point running Google Play Services and all the vendors bloatware background processes is just a waste of resources.


not really, there's postmarketOS. Although only a handful of devices work well, you can still get some use out of many phones.


Do we really need Samsung for that? I like the idea and I have some spare unlocked phones. It would be nice to have a webpage with ideas and downloadable code. Maybe something like a custom system image that would change an old phone into a simple portable arcade machine or something.


Perhaps like a Kubernetes cluster of old phones. A quick Google suggests others have looked at this before.

Edit: and I completely agree with you, we needn't have Samsung involved, as this doesn't have to be specific to their phones


Here's why, from the article:

> Friends inside the company told us that leadership wasn’t excited about a project that didn’t have a clear product tie-in or revenue plan.


Of course. Current market capitalism provides no incentive to reuse, upcycle, or recycle. Throwing a billion phones into landfills is obviously shitty for the planet and all of us who live on it, but hey it's basically free, so why do anything else? Unless we change our economic system, the only way this will ever get better is when the monetary cost of waste generation matches the real cost to us and the world.


Or i might be more likely to buy something that doesn’t turn into a brick in 3y.

Toyota and Honda didn’t get to where they are through planned obsolescence.


A high end security camera has a battery, can detect motion, make & store the video recordings (both locally and on a server) and it has a sim card in case wifi or ethernet is disconnected.

A video doorbell does something similar.


The funny thing is that Samsung is constantly throwing in free crap to try and get people to upgrade to their newer phones. If instead of yet another set of bluetooth headphones, they gave away free wireless doorbell enclosures or something, they could kill two birds with one stone.

Granted, I've never been accused of having much imagination so maybe I'm missing the plot.


This would be a huge undertaking for support/maintenance, not even counting in that due to the SoC any kind of security/Linux update is out of the question. Slapping on an enclosure, adding an app, and calling it a day would be a major security liability.


Interesting point, they would want the old device to live up to the name. Maybe refurb them into nameless crappy doorbells?


yeah, put a big samsung logo on it.

What would you [normally] charge if advertisement on your house was a market?


For a security camera, you need an IR array for night vision. Also cell phones draw too much power. (My cellular security camera battery lasts 3+ months).

Also I once tried to use my Pixel 1 as a off-grid hotspot. After ~4 days I would have to reboot it to restore connectivity (even though the hotspot was still enabled)


I want 30 cameras but I can afford maybe 2. The battery has to be good enough to continue recording for say an hour after the burglar switches off the power and unplugs the wifi.

The lack of night vision is a good point. Maybe a thermal vision adon? https://www.flir.eu/products/flir-one-gen-3/

Some battery led lights that switch on if the power goes out would also help.

I tried an app that turns the phone into a security camera but it just lost connection after a few hours. I suspect its just not a design goal.


So sad to see visionless executives at a company so large.


What an astonishing surprise!!

Anyone remember this - https://www.youtube.com/watch?v=6KCaS1z2MuE ?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: