lylecheatham

On the legalization front, Cato is definitely on the libertarian end of the spectrum of conservative think tanks, so that does line up

For context, this article is an opinion piece co-authored by:

Jeffrey A. Singer - A senior fellow at the Cato institute, a republican think tank that receives much of its funding from large republican donors/foundations and corporate donors. I don't know how to sum up the Cato Institute in 2 sentences unfortunately, but their wikipedia says plenty [0].

Josh Bloom - The Director of Chemical and Pharmaceutical Sciences The American Council on Science and Health, which is a pro industry advocacy group [1] that has received large amounts of money from the agriculture, petroleum, tobacco and pharmaceutical industries as per leaked funding documents in 2012 [2].

[0] https://en.wikipedia.org/wiki/Cato_Institute

[1] https://en.wikipedia.org/wiki/American_Council_on_Science_an...

[2] https://usrtk.org/industry-pr/american-council-on-science-an...

Except that the author of ruamel.yaml:

refuses to use git [0]

refuses to take community submissions (except through Stack Overflow? Seems like a misuse of SO) [1]

and refuses to implement .dumps() [2].

He is difficult to work with, and any time I need to debug code that intimately deals with ruamel.yaml types, I wince.

[0] https://github.com/pycontribs/ruyaml/issues/1

[1] https://yaml.readthedocs.io/en/latest/contributing.html

[2] https://stackoverflow.com/a/63179923/15170511

Ethernet usually has magnetic isolation through tiny transformers (oftentimes included inside the ethernet PHY IC) to avoid any issues with different ground potentials and such.

I still think Tile lost out on a gigantic market by not making it easier for hardware manufacturers to integrate into their ecosystem.

They should have been practically begging oems to embed it.

Aluminum Nitride actually has a favorable formation so they have to use much less friendly gasses like Sulfur Hexaflouride.

I mean most GIS software is pretty aware of precision models. Take GEOS for example, it's got a variety of ways to specify the precision in number of decimal places for a geometry. It still uses doubles underneath, but that's to avoid unneeded code complexity.

I really think #EndPatents is a very software oriented view of tech. In the physical engineering space patents are the only thing that allows a small company to actually design, manufacture and sell a product before a larger company can just squash them.

I know that getting investment as a small company in the hardware space would be near impossible without patents, because any investor without a brain would see that the giant in your industry could decide to take your idea, design it faster, manufacture it cheaper and sell it to a wider audience in a fraction of the time.

Yeah I think it's just how news media works I guess. Especially since 3D printing is a cool technology that hasn't quite hit it's stride yet, people are want to see how it's going to change things.

IM has been the same old IM for the past 5 decades more or less.

I'd say the real hype cycle for FDM was in 2013-2015, as evidenced by the SSYS stock price [1] (many other 3D printing companies have similar curves).

I think we're now on the tail end of the hype cycle graph and people are starting to find real uses for FDM

[1] https://www.google.com/search?q=ssys+stock (set to max time)

They are pretty fundamentally slow because of the constraints from the melt pool. Because you are trying to locally melt the material (lots of energy) and then immediately solidify it, you pretty much have to use a melt pool under a certain size and go under a certain speed to avoid making a total mess of the part.

Also a significant amount of energy goes to waste as it gets dissipated by the material, and the powder feedstock is extremely expensive as it needs to be a certain size of round powder which is hard to manufacture.

Minutes is a stretch. It will likely take multiple tools and setups to make a mold, and even more if you're starting from a billet that needs facing etc.

Apple actually gets a decent amount of coverage for their very impressive injection moldings. Most advancements in IM today are not in the ability to make complex parts, but more in the ability to make them pretty.

Apple makes many parts with extremely tight tolerances with almost no visual defects, and within the industry it's understood that it's extremely hard to make your IM parts look like Apple's for that reason.

Plastic scrap value is near zero [1] when compared to Aluminum

[1] https://webcache.googleusercontent.com/search?q=cache:fZ0f_O...

Yeah the plastic isn't recyclable, but e-waste is full of precious metals and structural metals like aluminum that can easily be recycled.

You'd be surprised actually! Prusa uses no gearing between the motor and the primary drive gear and it's a 200 step/rev NEMA 17.

You can transmit high frequencies quite easily, and I've actually seen bad stepper motors with high cogging torque transmit that periodic torque as an extrusion defect into walls with the correct thickness.

I'm pretty sure it's actually just a continuous under extrusion. It becomes periodic in nature because material gathers under the nozzle and then it hits the blob from the previous layer and gets pulled off. It's a similar effect to how droplets form and fall off of your faucet when the flow rate is low. This also leads to the diagonal angles seen as the Z-height goes up.

It's usually hard to get this to happen consistently, which they've done quite well.

I've got no citations, but I worked in FDM Additive Manufacturing for a few years, and spent 4 months of that designing extruders for a name brand company.

You're talking about a $40 tool just to flash the storage vs a part (micro SD adapter) that they give you with the SD card for free because it's so cheap.

Also doing a quick survey for the rated cycle counts on M.2 vs SD card slots:

M.2: I found this one [1] which is $0.768 for only 60 cycles

SD: This one [2] is $0.6256 for 5,000 cycles

I'm not sure why you'd say M.2 is more reliable, considering users often cycle storage dozens if not hundreds of times.

[1] https://www.digikey.com/product-detail/en/jae-electronics/SM...

[2] https://www.digikey.com/product-detail/en/gct/MEM2051-00-195...

If they didn't use SD cards, the storage would be more reliable, but users would spend a lot more time fixing bricked boards. By allowing for removeable storage (in a format that can be plugged into any other computer natively) they solved that problem because I can just re image the card and get going again.

The philosophy of the RPi is that they won't really add features to it unless a bulk of the user base would use the feature. For example they were hesitant to even build the WiFi into it because users who wanted that could always get a USB chipset, and building it in adds BOM cost.

Because with GSM you'd also need a plan for it, I don't really predict they'd add that. Especially since you can get it in hat format already.

People limping might serve as example that a 'state of the art, well trained neural net' can't achieve good motion with less than perfect hardware.

I think that people limping has more to do with avoiding pain than damaged 'hardware'. An example being people who are on large amounts of drugs being able to push through pain and further injuring themselves.

So how much of this is actually adressing the wrong problem

I think this paper addresses the right problem. By modeling the robot as a flexible system instead of a rigid system, performance improvements can be made in many scenarios.

Because there is no such thing as a perfectly rigid material (well at least within the realm of feasibility), even if the robot was designed to have extremely rigid and perfectly optimized joint angles and limb lengths, this technique would be beneficial.

Of course, where it really shines is when applied to a low cost, low weight system like the ones demonstrated in the paper. In the world of engineering, keeping things simple, low-cost and light opens many doors for using cheaper hardware and simplifying the design process.

If every time Disney wanted a new animatronic robot they had to get custom fabricated joints and limbs, the costs would be exorbitant. If instead they could just reach into their standard limbs box and slap it together, and then let the software fix it they save money and effort.

Desktop Metal | Embedded Software Engineer | Boston | Full-Time

https://www.desktopmetal.com

Desktop Metal is revolutionizing 3D printing of metal parts. Our printer’s robust build processes create complex parts beautifully at a price attainable by any design and manufacturing team and can be used within an office environment.

Embedded Software Engineer: https://boards.greenhouse.io/desktopmetal/jobs/620949

It says senior, but I know we are also looking for junior engineers on the team as well. Also it's not your plain-old-embedded development, the code you write makes the machine move so it brings in plenty of math and physics.

They're actually using servo motors! They refer to it as a "servo-driven parallel gantry system" in the abstract.

Checkout the second to last page for pictures: https://arxiv.org/abs/1709.05918

I think you bring up a good point about the lack of precision at high speed. I think the next big thing in this industry is going to be diving into the factors behind the 'sloppyness' you mention.

I'd like to know how well the auger is able to do this.

I can answer this one! it's actually more like a nut on a bolt, instead of an auger, meaning it can reverse just fine.

Although, I disagree that a screw drive is needed. I use hobbed gearwheel drive on my extruders and rarely is the issue the hobb slipping. Typically, it is the motor stalling instead.

Although that may be the problem on your printer, it's simple to fix because one can always add a larger motor. The problem that they solved is that eventually, a stronger motor won't cut it because the hobbed gear will slip, and that's what this paper is all about. Optimizing the motor is done by just upscaling it, but optimizing the drive is what made them to move away from the existing hobbed gear setup.

They could go faster if they used a 32 bit controller like a smoothieboard or duet instead of the 8bit "arduino based" controllers. {Not saying RAMBO is a bad board, they are known to be very reliable for OEMs}

I would actually predict a minimal increase in speed. 32-bit controllers don't make for inherently fast printers, they allow for it if utilized properly. What you'll find though is that Marlin and Smoothie both use the GRBL planner under the hood, and generate very similar output. The only major difference I noticed in the output between the two is that Smoothie generates smoother (hence the name) step waveforms.

I guess you can also generate steps at a higher frequency with smoothie, but you should be careful equating higher step frequencies with higher speeds, when most of the time people are running with microstepping settings. Higher frequency allows you to get more dynamic range between speed and precision, but that's not really what the paper was about.

It's actually just a standard 4-40 thread cut into the filament. They use rollers to prevent rotation of the thread when cutting the threads into the filament, and then when they actually use the nut it's got enough thread engagement that it's better than a hobbed gear. If you think about it, it's almost like a hobbed gear that attacks from all sides instead of one or two.

Feel free to ask me questions on it, I work in the industry and have spoken with one of the authors of the paper, it's a shame it's behind a paywall.

I wouldn't compare it to the screw drive they use on injection molding. Those are for pushing pellets and melted plastic, not for pushing a rod (or strand) of continuous plastic.

This design uses threads on the outside of the filament [1] to grip it with a nut.

The main difference is that the design and manufacture of this system is extremely simple, contrary to what you've said. They use off the shelf 4-40 [2] taps and nuts to create a triangular thread. I would even go as far as to say this is _easier_ to manufacture than hobbed gears.

[1] http://www.sciencedirect.com/science/article/pii/S2214860416... Excerpt for those without access: "in the FastFFF system, this limit is overcome using a rotating nut to drive a threaded filament (Fig. 2c), analogous to a leadscrew linear actuator"

[2] http://www.sciencedirect.com/science/article/pii/S2214860416... Excerpt for those without access: "The ABS filament is cut with a standard 4-40 thread (Craftech Ind.), which has a triangular profile (Fig. S1, Table S1) [10]."