Gov’t Contractor Copies Open Source 3D Printing Concept and Patents It | Hacker News Reader

Gov’t Contractor Copies Open Source 3D Printing Concept and Patents It

techdirt.com

233 pointsdanboarder83 comments

While most of the conversations below talk about specifics of the claims ... I think there are much larger issues at play of which this is a prime example.

Innovation is a series of steps. The major problem is that you can Open Source a bunch of majorly innovative things, to only have some very smart people that have studied in the domain write down the logical next steps and have your entire area of amateur research now blocked by patents. They are free to acknowledge your work but now say, "xyz is new innovation in this direction". There is no LGPL/GPL for patents. Unless you are constantly trying to predict the next step, capital (and in this case government, see below) will be there to take innovation private. I see this as a massive theft vector of the commons. It allows private capital to front run public Open Source innovation. Yoink.

There are two other perverse aspects to this. In academia amateur research isn't really citeable in the same way, to the point where it isn't even acknowledged unless the work is so profound.

The third one, that most likely applies here, is the Bayh-Dole Act [1] that allows government money to make for profit patents on research.

*edit, 4th one.

The patent office is a profit center. And sorting out the legitimacy of patents is left to the courts. Patents aren't tools of innovation, but legal instruments that protect corporate oligarchies.

[1] https://en.wikipedia.org/wiki/Bayh%E2%80%93Dole_Act

This is why I think that the patent examination process should include a challenge-period, where anybody can suggest solutions to the problem posed in the patent-application.

This already exists.

How does one challenge a patent?

How to challenge a patent application that hasn't been granted yet based on prior art:

https://www.uspto.gov/patents/initiatives/third-party-preiss...

Note that I'm not aware of any way to challenge an application on non-prior-art grounds like 101, 112, etc. But I imagine emailing the examiner assigned to the application would work. Unfortunately examiners aren't assigned until a few weeks to a few months before examination, so a challenger will need to periodically check the status of the application before contacting the examiner.

You can look up the examiner's phone number here: https://portal.uspto.gov/EmployeeSearch/

Email addresses are typically firstname.lastname@uspto.gov.

This is something else. Here, not the patent itself is challenged, but the work of the patent officer is made easier because prior art is supplied.

What I propose is that anyone can submit solutions to the original problem stated in the patent, without yet knowing the solution proposed in the patent. This allows us to demonstrate that the solution is trivial to those who are skilled in the art.

Unfortunately the US doesn't work on that legal standard, though I think that's a nice approach.

In practice, almost no third parties submit prior art, which is a lot simpler than submitting solutions. So I doubt people will submit any significant amount of solutions.

If there was a forum like HN where every post was a patent-problem, I would be happy to participate in the discussions about possible solutions, on a daily basis.

Until the legal standards change, look into submitting prior art. The examiner and public will appreciate it.

This isn't copyright law where independent authorship/creation is a defense. The US is a first to file nation, which it shares with nearly every other major country in the world.

>How does one challenge a patent?

With massive amounts of money, since the PTO has made it impossible for the system to actually be fair.

This is the case: the whole pre-issuance period a counter party can challenge and provide invalidating arguments. Nine months after issuance there is also a challenging period. Additionally you can invalidate a patent by litigation but this gets more expensive.

Yet many bogus patents still slip through. Challenging patents is a costly and time intensive process.

This is different; this is not about finding arguments why a patent-application should not get through (i.e. the "search" part of the job of the patent officer). This is about finding the solution that is proposed in the application, which shows that the solution is trivial to those skilled in the art.

speaking of costly, their gofundme[1] says "We need $60,000 to pay for attorney fees and patent office fees related to a re-examination of the patent application."

That's quite a bit for an opensource developer to have on hand, but peanuts for major corporations.

[1] https://www.gofundme.com/f/keep-hangprinter-free

or better: simply abolish the patent system.

Or, keep the system, but allow the PTO, and individual examiners to be liable for any fees resulting from an invalid patent being granted in the first place.

Individuals (like the original project in this case) and the public at large (court/challenge costs) shouldn’t be burdened with the costs of fixing the PTO’s fuckups.

I am sure so many examiners would be willing to take that job.

Or take this as an incentive to actually do their jobs properly.

All my own personal opinions. I work in IPR.

You're basically closing down the patent system here, any prior disclosure anywhere in the World, in any language can be used to show the invention lacks novelty.

There are some restrictions in different jurisdictions on what can be used to show there is no invention (ie there is no 'inventive step' or it's 'obvious' [it's a legal definition, not just a dictionary definition of this term that is used]).

If you're going to hold an examiner liable (there's actually a law in the UK protecting civil servants from liability, but hey) then you're going to have to give them lots of time and lots of resources ... that's going to make it all too costly.

You'll spend more than is profited from most patents in order to be sure that a patent won't be challenged in court ... and it still could be, just under other facets of the law, such as whether the disclosure is sufficient to work the invention.

You have to allow people to be less than perfect. If you're a programmer, imagine you had essentially unlimited liability if you created a single bug (examiners don't know if the patent is worth nothing, or 100s-of-$Millions, that's all down to externalities).

Could the system be improved, for sure, but it's always going to have to balanced against cost of administration. If you're court system is working effectively you don't need anything like a perfect parent examining system. If you're courts aren't working ... well fix them instead of spending disproportionately elsewhere.

All good points; I will however point out something about this specifically:

>then you're going to have to give them lots of time and lots of resources ... that's going to make it all too costly.

If the system has to become prohibitively costly to get it to a level of semblance of actually working, then one has to wonder if it's worth it at all for it to exist.

The patent system works currently...

...somewhat

It doesn't matter how expensive your telescope is, if you want it to have perfect optics you'll have to spend more. You need to decide at what point it is good enough. If your post-processing is good (courts) then you can cut costs on the optics (patent examination) and still achieve the results you want.

But I wouldn't expect my neighbour to have to pay to fix/improve the optics on my telescope.

One of the major parts of the obviousness tests in the US (and in every other major jurisdiction) is whether a person skilled in the art would even have understood the problem, recognized there was a problem, etc.

I don't think challenging patents is the whole answer, all the things I outlined operate within the current patent system. The rules are broken. They only way that Open Source has to currently prevent this taking from the commons is

1) Aggressively innovate and publish, including using the same front running techniques as used by industry and academia.

2) Aggressively challenge all patents in a classification. Rapid prototyping and manufacturing, robotics, farming and transportation are all huge land grabs that are going on right now.

3) Get the Bayh–Dole Act repealed or neutered. A penny of public funds (including the facilities, any of the staff, etc) and it should be a public good, available to all.

4) Make the patent office no longer a profit center

Why have a patent system when it doesn't even take any responsibility for ensuring the patent is valid? Whether or not this patent is allowed to remain depends entirely on how much money the open source guys would be able to put up to take it to court.

Ultimately because we have a system in place that doesn't care if people get exploited. Our American bureaucracy is designed to exploit, lie and cheat. Governments protect wealth not people.

And the wealth does not want new, does not want change, does not want disruptions. Patents are not there to protect innovation capability, they are there to discourage innovation and "chill" creators.

In another HN article today, there's a Harper's Magazine article about the gullibility of people and online scams. Having written and acquired a global patent, to significant personal and financial expense, and as a result have been exposed to the industry of scans surrounding patents and patent law, I now feel the entire patents enterprise is a giant multi-industry scam, institutionalized by major corporations in concert with their governments because that scam works for them very, very well. It is institutionalized now, and no longer called a scam officially, because it is an official conduit for stealing valuable working ideas from the non-1%.

This is the claim they were granted:

“A cable-driven additive manufacturing system comprising:

an end effector configured for linear translation within a three-dimensional workspace along an x-axis, a y-axis, and a z-axis;

an aerial hoist suspending the end effector within the three-dimensional workspace by at least one suspension cable, the aerial hoist having a static position within the three-dimensional workspace whereby the aerial hoist is not translated along the x-axis with the end effector and the aerial hoist is not translated along the y-axis with the end effector;

an aerial cable winder configured to control linear translation of the end effector along the z-axis;

a plurality of motion-control base stations disposed below the aerial hoist and configured to control linear translation of the end effector in an x-y plane defined by the x-axis and the y-axis, each motion-control base station comprising a base station cable winder;

a plurality of motion-control cables, at least one of the plurality of motion-control cables running from each of the base station cable winders to the end effector;

at least one tension-control base station; and

at least one tension-control cable running from the tension-control base station to the end effector.”

Considering all the requirements, it is quite narrow. This means if a particular implementation doesn’t have all of these things it is not covered by this patent.

Don’t let logic get in the way of anger, that’s not how the internet works these days.

It's not about the number of requirements. What you quoted is actually a general description of the device.

What it says is you have an effector (effector is a generic word that means something that does something) suspended in space. A rigid hoist provides a fixed position above the work area to suspend the effector. The effector hangs from the hoist by at least one cable. That cable is used to control the z postion of the effector. There are also at least two other controllers used to control the x and y position. Each controller is connected by at least one cable which is used to control position in x and y.

In fact this patent was probably granted because that clause is not specific to 3D printing. The "effector" could be something else. Like a lathe or drill or camera or lightbulb or anything that does anything.

> The "effector" could be something else. Like a lathe or drill or camera or lightbulb or anything that does anything.

This is just an algebraic transformation on something specific to something generic that supplies +-energy and/or +- mass. Running a macro over someone elses idea shouldn't allow one to patent something.

Such robots exist since ever. Here a 12 year old youtube video.

https://m.youtube.com/watch?v=RCa8uDFzbsw

My comment was around mechanical transformations of ideas to make them more generic and then patenting that generic "discovery". It had nothing to do with the specifics of cable robots.

Ooh, that's an interesting point, though I'd see it as even weaker as a result. It isn't even to the level of "What happens if I put a 3d printer pen on a Robocrane?", but stops at "What happens if I put something on a Robocrane?".

I might not be interpreting this correctly; can you unpack your argument for why this wouldn't cover the HangPrinter? Is it because the motion-control base stations disposed below the aerial hoist can control linear translation of the end-effector not only in the x-y plane defined by the x-axis and the y-axis, but also in the z-axis? Do patent claims work like that?

I don't see how you could build a cable-suspended 3-D printer without including all of these elements.

As usual, the media doesn't read the claims, which are the legally enforceable part. To those without a background in patent law, how similar the abstract, drawings, or parts of the specifications are to prior art doesn't matter in terms of granting a patent. What matters is the claims. Let's figure out precisely what the difference is between the claims and the prior art that lead the examiner to allow the application.

You can take a look at the patent application prosecution history here: https://patentcenter.uspto.gov/applications/16383105/ifw/doc...

The applicant amended the independent claim to add the following:

the aerial hoist having a static position within the three-dimensional workspace whereby the aerial hoist is not translated along the x-axis with the end effector and the aerial hoist is not translated along the y-axis with the end effector

In the "Notice of Allowance and Fees Due (PTOL-85)" the examiner discusses why the prior art doesn't have this feature. I don't have time to read the notice, but this is the longest notice of allowance I can recall. The examiner seems to be going into detail about why they believe it is different.

The Hangprinter was submitted by the applicant as prior art and considered by the examiner: see "List of References cited by applicant and considered by examiner".

Note: I'm not saying the examiner was right to grant a patent here, just that critics of a granted patent should start with the claims as that's the legally enforceable part (and not the drawings or abstract). And typically the examiner's logic in allowing the application is documented, so critics should take a look at that too.

At least if the hang printer is acknowledged in this patent as prior art, but sufficiently different to justify this patent , isn't that basically tacit acknowledgement that hang printer would not be infringing on this patent?

Not a lawyer, just pondering...

To me, this doesn't seem like a large number of claims. It seems like the minimum number of claims that appear as a result of attaching a 3d printer head to a cable-driven parallel manipulator.

> A cable-driven additive manufacturing system comprising:

So, it's a 3-d printer with something additional for it.

> an end effector configured for linear translation within a three-dimensional workspace along an x-axis, a y-axis, and a z-axis;

Nothing new here, as this is true for any 3-d printer with a stationary bed.

> an aerial hoist suspending the end effector within the three-dimensional workspace by at least one suspension cable, the aerial hoist having a static position within the three-dimensional workspace whereby the aerial hoist is not translated along the x-axis with the end effector and the aerial hoist is not translated along the y-axis with the end effector;

The printer is suspended, and the suspension point doesn't move. So, distinct from having the printer bed move, but otherwise the obvious next step if you have the head moving in 3d.

> an aerial cable winder configured to control linear translation of the end effector along the z-axis;

Again, the obvious next step. With a fixed suspension point, pulling on the rope sets the height of the printer head.

> a plurality of motion-control base stations disposed below the aerial hoist and configured to control linear translation of the end effector in an x-y plane defined by the x-axis and the y-axis, each motion-control base station comprising a base station cable winder;

If x-y control is handled from above, then the maximum x-y speed is limited by the speed of gravity. If you attempt to move any faster, then one of the suspension cables will no longer be under tension, and the remaining ones will act as a pendulum until the maximum length of the untensioned cable is reached.

> a plurality of motion-control cables, at least one of the plurality of motion-control cables running from each of the base station cable winders to the end effector;

Well, yes. Otherwise the base stations aren't exactly base stations.

> at least one tension-control base station; and

If you are only controlling the length, then any allowed tolerance in the length spooled out will result in a larger uncertainty of the location of the print head. This is a requirement from having the x-y controls below the printer head.

> at least one tension-control cable running from the tension-control base station to the end effector.

Well, yes. Otherwise the base stations aren't exactly base stations.

So these entire claims boil down to nothing more than the shower thought of "What would happen if I put a 3d printer pen [0] on a Robocrane? [1]" Given cable-actuated parallel manipulators aren't exactly new [2], this wouldn't pass the non-obvious test to me. Heck, applying this for 3d printing is suggested in this reddit thread [3] (2016 video, March 2021 thread).

[0] https://www.amazon.com/3d-printing-pen/s?k=3d+printing+pen

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

[2] https://journals.sagepub.com/doi/10.1155/2014/823028

[3] https://www.reddit.com/r/robotics/comments/mej19w/extremely_...

> Considering all the requirements, it is quite narrow.

This is clearly not narrow. Think of the last system you designed - any system. Would you be able to fully describe it in 7 bullet points? Or would those bullet points have to be exceptionally open-ended, such that they would capture your system and many, many others?

Patent trolling is the worst kind of parasitism. Patent laws must be changed to penalize harshly this kind of behavior (after due process) to disincentivise people who choose this path.

Just thinking outside the box for a second, but shouldnt it be possible to sue the patent office for copyright infringement in this case?

As in, they published someone elses copyrighted design without giving appropriate credit.

Are they given special exemptions from copyright law?

DMCA takedown notice almost seems appropriate.

Most likely you can't because of "qualified immunity" wherein government officials can violate rights of people and face no consequences.

Often bc the alternative is nothing ever can get done. Which maybe if you think there should be no systems at all you'll think is a good thing, but the majority of people enjoy some things the government provides.

The alternative is that officials face consequences for violating peoples rights. This would actually make things work better because rights would be enforced. Officials who violate them would be fired and/or fined and/or imprisoned (depending the situation), because what they did was illegal.

You can't have judges potentially getting punished for making mistakes. There are a lot of examples. No one would take those jobs in your world.

People get punished for making mistakes in every other type of job and people still take them, so why should judges get a free pass?

Are you saying that you'd be OK with being personally responsible for a bug you put into your code that causes some harm down the line? As in, costing you $$$$$$$$ or jail time?

That is what it sounds like you're saying.

This is already the case. If wrote code for say a heart implant device and I did a shitty job and a bunch of people died because of it, I for sure would be fired, and likely sued, and possibly even face jail time (depending on the specifics). So yeah, that's the world real world for most of us.

Wouldn't that be the applicant's fault rather than the patent office? The patent office reviews the application but is not responsible for its contents. They publish granted applications as a matter of course. But they cannot be omniscient about knowing when a bad application comes through; they can review to their greatest ability but will inevitably not always be right.

My thought was it should/could work like say with youtube:

https://support.google.com/youtube/answer/2807622?hl=en

https://donotpay.com/learn/youtube-dmca/

All that should apply to the patent office publishing material as much as anything else.

When you submit your patent application, typically the text and drawings become public domain.

Looks like the “company” that filed for the patent is basically the corporate research arm of UT.

https://en.m.wikipedia.org/wiki/UT–Battelle

This makes the situation even shittier but also makes it more likely that the people who were issued the patent had little choice in the matter as it’s quite common for universities to basically patent every crap that a group of grads and their advising prof. produce on campus.

Do the actual inventors have to go through the parent litigation process, why not report it as theft to the local authorities, and use that theft report as a pass to keep working. Why is it for patents that the burden of proof is inverted?

Look up the difference between criminal and civil law.

Civil crimes are for the rich and criminal are reserved for the poor.

That is in a sense true. If you don't have any money to begin with, you don't have much to lose by getting sued and there is not much incentive for anyone to bother.

Personally I think it's better for everyone that the police are not usually involved in the enforcement of IP law. A lot of corporations would love the state (aka taxpayers) to pick up more of the tab for keeping mickey mouse out of the hands of the public.

The flip side of what the parent is asking for (the police treat "theft of ideas" like criminal theft, without even any registration of the IP!) would very likely result in clogged courts and mass incarceration of downloaders.

(Admittedly this ridiculous hypothetical would also require that police have time and resource to treat theft of property seriously, which is not true in many places and considered no big deal given that you would most likely not be compensated for your losses except by insurance in any case).

Very simple solution, just ignore the patent and set anyone who tries to enforce it on fire

That sounds facetious. I would try reaching out to the thieves themselves to discuss the prior art, or possibly the EFF:

https://torbjornludvigsen.com/blog/hangprinter-is-prior-art/

https://www.eff.org/issues/patent-busting-project

lindrf@ornl.gov

postbk@ornl.gov

lovelj@ornl.gov

chesserpc@ornl.gov

roschliac@ornl.gov

Next thing you'll tell me to vote in November. Thanks but I like my plan better

Please don't advocate for violence, even if it's only in jest.

Why not? If you want someone to (not) do something, you need to provide a compelling rationale. Otherwise you're trying to compel some based on your personal preference.

Personally I think wild hyperbole is funny and effective at making a point, so my preference is that the poster continue.

> If you want someone to (not) do something, you need to provide a compelling rationale.

OK. Advocating violence is a crime. And text-based communication is notoriously difficult medium to express sarcasm and irony. Crimes are bad, mmmmkay?

Nitpick - advocating violence, specially as a personal vida is not a crime in most countries; promoting violence is, and it depends on the violence -death penalties and stonings are state-sponsored acts of violence, perfectly legal within their applicability in many parts of the world. Also, crimes arent necessarily "bad", nor requinte violence. abortion in some American states is a crime, and stealing cars is also a crime.

> Advocating violence is a crime

The Dunning-Kruger effect is real.

You're probably thinking of assault. Assault requires an imminent threat of physical harm and the apparent ability to inflict the harm.

Mike Tyson threatening to punch someone in a bar is assault. Saying that you're going to set all patent trolls on fire is not, unless you have Human Torch like superpowers.

Advocating violence is not necessarily a crime. For example, in the US that violence must be imminent and likely (see Brandenburg v. Ohio).

To me, calls for violence are exactly the dividing line between free speech and restricted speech. Show extreme judgement when invoking violent imagery.

The EFF (Electronic Frontier Foundation) might be able to help. They regularly take on and challenge bad patents.

I can't speak to the issue of whether this particular patent is valid, nor whether UT-Battelle plans to enforce it or how vigorously (some patents invented by DOE contractors are licensed quite reasonably. I know this because I was an inventor on a few of them). If the patent turns out to be invalid one could file a protest with the Department of Energy.

Further, it looks like this contract will be re-bid in 2024 [0]. So if one felt strongly that UT-Battelle's behavior here was improper one could protest to DOE any extension of the contract to UT-Battelle. I rather doubt that a single bad patent would make much difference in DOE's decision, but you never know.

[0] https://primecontract.ornl.gov/

What proof do they have that it was copied, other than "it looks similar to ours?" Multiple discovery happens all the time, and the patent process - for better or worse - is an effort for an inventor to officially put their "I made this" stamp on.

I'm not sure if you're (just) playing devil's advocate, but the blog post[0] linked in the article covers quite a lot of instances of prior art[a] / multiple discovery / coincidences, going back years before the patent application was filed. Whether this is proof that the design was copied -- I can't say.

[0] https://torbjornludvigsen.com/blog/hangprinter-is-prior-art/

[a] what I would call "prior art", anyway. IMHO, IANAL (AFAIK)

Proving that it was copied is only necessary to show copyright infringement. For example, if you independently write a Book A1 that happens to be very similar to an existing Book B, it is not copyright infringement. If you write Book A2 by taking an existing Book B and changing the name of each character, then it would be copyright infringement. Even if A1 and A2 end up having the exact same words as each other, the origin of those words matters.

For a patent to be granted, the implied argument is entirely different. For a patent to be granted, it must be both novel and non-obvious to somebody in the field. If a patent is granted that exactly describes something that already exists, then one of two things must be the case. If the patent description was derived from the existing work, it wasn't novel. If the patent description was derived independently, then it was the obvious solution to a problem. In either case, it fails at least one of the requirements for a patent.

Unfortunately, the patent office has frequently interpreted "novel" as "not covered by a previous patent", with the implication that every invention must have been patented at some point. The courts can go back and invalidate a patent, but that would be a defense against an actual lawsuit, and would carry significant risk to initiate.

IANAL, but in this case there is so much prior art that I would just ignore the patent. If someone comes trolling, then that would be the time to challenge the patent. You might even be able to so it on the cheap with an ex parte reexamination.

Sure if you have 7 figures to spend on it in court, and the patent troll might still be able to get injections especially if the printers or parts that “violated” the patent in question are imported.

Ex parte rexamination should be far cheaper than that. The success rate is lower than other avenues, but could work due to the mountain of evidence.

how can you patent something with prior, public art?

One possible explanation… Patents go to the first to file. The prior art search focuses on prior patents. Patent examiners can’t consider something prior art until they know about it. It is possible to invalidate a patent after the fact in situations like this, but that’s expensive (as the article describes).

Info on first to file:

> The first inventor to file (FITF) provision of the America Invents Act transitions the U.S. to a first-inventor-to-file system from a first-to-invent system and became effective on March 16, 2013. The provision introduced changes to 35 U.S.C. § 102 that impact patent prosecution directly. https://www.uspto.gov/patents/first-inventor-file-fitf-resou...

I didn’t click through to look at the actual patent, but it’s also possible the patent was narrowly written to focus on an industry specific application.

This is only partially true. When you file a patent, you have a 'duty of candor.' The filer _must_ disclose all prior art they know of. If they don't, it's straightforward though time consuming to invalidate the patent.

We need something similar to the EFF but for invalidating patents and other IP protections given to those who got them through the design of the lawfare system. It's too expensive for most small to medium size companies to take on a multinational that has fooled the government into granting them an exclusive license over some part of innovation that they didn't actually innovate.

EFF is actually one of the orgs doing some work in this space. Another great reason to be a donor! https://www.eff.org/pages/defend-innovation

You can patent anything (https://patents.google.com/patent/US6368227B1/en) and the barrier to having it granted is incredibly low. The patent office works on the premise of granting the patent and letting the lawyers figure it out afterwards. Enormous volumes of nonsense patents are granted every year, "inventors" who then go forth and try to remora off actually innovators.

Reply on news.ycombinator.com