Bootstrapping a New Technology? 360
djk1024 writes "I've just filed for a patent on a new approach to motion capture that is simple, cheap, easy, accurate, and portable. It's RF-based, accurate to 1 mm, and simple enough that a sophisticated hobbyist could build one in a couple weekends from plans and standard electronics. So now what? I quit my job and have been working on this full-time for the past couple of years; now I'm out of money so can't continue development on my own. I'm also not an electrical or RF guy so I can't carry out my own independent development on the electronics. I'm quite frustrated at this point. I've been in the software development field for over 30 years and have gone through a large number of startups, but always just as the head techie, and always as part of a team. This doing it alone sucks. I would love some advice on how best to go forward."
Re:Sell your patent (Score:5, Informative)
Now, I have one question and please don't take this the wrong way: if you system is so simple it would take only a few weekends to build yourself, why is it taking you so long to develop?
Development usually takes a lot longer than following somebody else's directions.
.... If you just filed for a patent you should be patent-pending soon
If he just filed a patent application, then his thing is "patent pending." As soon as you file, you can call it that.
which means even if you sell kits or samples the users of your kits/samples will not be able to mass produce all they want - patent pending alone is enough to bring up and win a court case.
Um, no, that is absolutely not true. You cannot sue on a patent until it issues. Before it issues, you don't even know what the claims are going to be when they issue. In most cases, they get amended during prosecution. So until your patent issues, you can't sue anybody, much less win. There is one thing to be aware of. Once your application is published, if you put a potential infringer on notice of your pending application and if it then issues with substantially the same claims as the ones that were publishes, then after it issues and you sue them, you will be able to get a reasonable royalty going back to when you put them on notice. But you still can't actually sue until the patent issues.
Also, you don't really need a patent to copyright or license your idea, so why not do that now?
You can't copyright an idea. You can copyright your description of it, but that doesn't prevent somebody from reading that description and implementing the same idea. It just keeps them from copying your description. Copyrights and patents are not interchangeable.
I'm a patent attorney, but this post is not legal advice. It's for entertainment purposes only. In other words, if you use a post on Slashdot as legal advice and things go badly for you, (1) you deserve whatever you get, and (2) don't try to sue me.
Re:Patent (Score:1, Informative)
I know you're just joking, but...
Patents don't preclude hobbyists from building their own, so long as they don't go selling it. There are a lot of patented technologies out there, that us hobbyists casually build or modify all the time, without a problem. Most of the time it's also OK to sell that stuff, so long as it's just the hobbyist selling personal belongings that they no longer need. (That is, not mass producing for profit.)
The whole point (or it was intended as so) of patents is that the methodologies are out in the public, which allows everyone to know the method and even build on it, while allowing the inventor to be able to be the sole producer of it for profit purposes. If you don't want the public to know about it, you keep it as a trade secret. No one (might) know about the methodology (security through obscurity?) but you have no protection either.
Filed way too late. (Score:3, Informative)
You've spent two years on this, and only now getting into the patent? You should be prepared for disappointment. The odds are very high that someone has already patented your exact idea, especially if it's something that you could figure out with no prior RF electronics experience. It's difficult to let go of something that you've put your heart into for so long, but you need to be prepared to do that if you run into problems. Otherwise it could drag you even further down.
Re:Great idea (Score:5, Informative)
Sorry to reply to my own post, but I may have been too quick on the draw to reply. I'm sitting here thinking he wants to take over Hollywood when maybe all he's after is an inexpensive personal one-on-one mocap system that small companies could use to capture one or two people. In that case, he can throw what I said above out the window.
I apologize for my knee-jerk reaction.
What a douche (Score:2, Informative)
So you've got a "great idea" about how to do this but you're "not an RF guy". $50 says your "great idea" is one of the following:
Note that even trying to find your application turns up a number of older apps (2007 and 2004) that seem to have the same idea.
Re:Sell your patent (Score:4, Informative)
I could create software, hardware, an image, a document, an audio recording, it doesn't matter as soon as I create it and release it somehow I automatically get my rights to it. If someone were to copy it, regardless of weather or not I held a patent, I could raise legal action against them (unless I had already released it under a particular license which granted them use).
Yes, absolutely. But that still doesn't get you where you need to be. For example, let's say that I invent a brilliant new circuit, and do up a nice schematic of it. I have a copyright in that schematic. Maybe I'll even register my copyright. And if you copy the schematic, you infringe my copyright. But if you get a copy of my schematic and build the actual circuit, you have not infringed my copyright. The only way I can keep you from building the circuit is by patenting the circuit.
The company I am a part of provides such development services as well, and I'm fairly confident if the author claims his system is as simple as he states it is we could have developed it for him into a product for less than $5,000US and in less than a few months.
You ought to send me your contact info. I sometimes have inventors who need that kind of service. It's nice for them to have options. Contact info. [jw.com]
Re:Sell your patent (Score:3, Informative)
Re:Sell your patent (Score:3, Informative)
Read the title of the parent's post... he's suggesting to hypothetically sell the patent, and the hypothetical buyer.
Re:One question (Score:5, Informative)
Re:Patent (Score:4, Informative)
I know you're just joking, but...
Patents don't preclude hobbyists from building their own, so long as they don't go selling it.
That is false. The statute makes it illegal for anyone other than the patent holder to "make" the invention. Selling it is also illegal. See 35 USC 271(a)
Re:Filed way too late. (Score:3, Informative)
Re:Patent (Score:4, Informative)
That is then another US extension of the original idea behind IP. Patents are supposed to contain a full disclosure of the invention, so that anyone using just the patent can build one, legally. Then this person can have a look at the invention, and use it as inspiration to improve on it, and potentially patent the improvements. That is the idea at least - that is how a patent stimulates invention and technological progress.
Whether such an improvement infringes on the original patent (and thus needs a license) will vary: if it uses the patented technology directly with minor modifications (e.g. an addition) it probably does, if it uses the idea but implements it in a different way, then it probably doesn't. When the patent expires of course the invention ends up in the public domain and can be used by anyone. Which is again one of the purposes of the patent system.
Selling a product based on a patent that you do not own or have a license on, is indeed illegal. Though reselling such a product is not. I.e. I invent something, patent it, make a product, and sell it to you, then you are free to sell this product to someone else. I can not limit that anymore.
Re:Contradictory? (Score:4, Informative)
Re:Help will be required (Score:3, Informative)
Re:Open source? (Score:2, Informative)
Re:Sell your patent (Score:2, Informative)
Next Steps... (Score:2, Informative)
If you would provide an email address, or another way to get in touch with you -- outside this public forum -- I can give you pointers on how to go about this. I have a lot of experience in this area. You are in a unique position in that you have an idea that you believe works -- based on prototypes -- but you lack a business case and at the moment a complete product. What you need to do is basically show people that this is not just an idea and that it is feasible to move forward with this. I understand that you've built a prototype and have a patent but that it is still not engineering proven -- which makes it risky.
There are ways of further developing this, and marketing it to other companies as IP or as a dev kit with licensing/integration option.
You have to dot the i's and cross the tee's before someone will look at it as a viable technology. Doing so may not be as difficult as it seems depending on what you've done so far. To use a familiar analogy: think of it as having built an engine but only having a command line front-end, wrapping a GUI around it may not be technically difficult but it has to be done correctly from a UI perspective for your first demo to wow the user.
You are in a very good position since you've applied for a patent. That makes it easier to talk to other people. I'd also like to remind you to always get an NDA before talking in detail to anyone.
As I said, I'd like to get in touch with you to discuss this further or at least take up an hour or two of your time in exchange for $0.02 of advice.
-Arvin
Re:build a working sensor using the information gi (Score:3, Informative)
And this is exactly why I don't thing software patents make any sense.
Part of the problem is that it does not cover an implementation (i.e. a solution to a problem, e.g. a paper clip is an implementation of a solution to keep two pieces of paper temporarily together without damaging them, but there are other ways to do so). Software patents cover only the idea (in this case "we can hold two pieces of paper together without damaging them"). A paperclip is not copyrightable on the other hand. Everyone could copy the paperclip but only sell it freely after the patent on this invention expired. That is a long time ago so now everyone can make paperclips.
Patents are supposed to cover solutions, not problems. An algorithm may be such a solution (e.g. an algorithm to compress audio with relative little loss: the mp3 compression), but then that is a mathematical algorithm and as such not patentable (at least under the original patent agreements - at least in the US it is patentable). This I would seriously call a matter of debate whether such an algorithm should be patentable, as it comes quite close to a technical device. The actual implementation of such a compression algorithm would fall under copyright again. So in the US, mp3 compression is patented, and on top of that implementations are copyrighted. So to use an implementation one would have to get a patent license and a copyright license (and possibly pay for that). Without the patent it would be OK to figure out how the software works (reverse engineering, another thing that was explicitly allowed and even encouraged under the original patent agreements), and then create a separate implementation by yourself. Then you create a new implementation, you have the copyright, but you may have to pay patents still nowadays in the US.
Patents, also software patents, do not contain complete source code as far as I know. These patent an idea (the "single click sales" patent to name an infamous one), just an idea, not an implementation of the idea. Even when a software patent expires there will not be any source code that suddenly appears in the public domain as code falls under copyright.
When open sourcing your software it still falls under copyright. Even if MS would open source Windows and publish the source code, that would not necessarily mean you can copy it and give it to someone else. They can tell you what you can do with it and what not. MS still has the copyright. The Linux kernel is open source and also copyrighted: by many different people, all covered by the GPL allowing copying and redistribution under certain conditions.
Re:Sell your patent (Score:3, Informative)
Labeling something "patent-pending" does nothing other than make you feel good. It has no legal effect at all.
That's not true. If you release it and don't tell someone it's patent pending, they can copy it and won't have to pay royalties till the patent comes out. If you put patent-pending on it, you can go to them for retrospective royalties once your patent comes out.
N.B. Exact details a bit fuzzy. I'm not sure if they count as "willful" before the patent issues or not. They certainly will be counted as "willfully" using the patent afterwards.
Re:Sell your patent (Score:3, Informative)
This is why we don't start our posts (Score:5, Informative)
in the title
Re:Contradictory? (Score:3, Informative)
Because I had an expert in RF build my prototype and those were his measurements.
Most of the posts are about "MBA business plan" type stuff, I'll try to go another direction. You need to get many hours of engineer time via consultation or hire, a RF EE and a machine designer or line designer ME or industrial engineer.
Ah, so a prototype was built and it worked. That is only about 1/10 of the EEs job. Actually that sounds like something an EE would delegate to a tech. Now you need an RF guy whom can deal with UL listing, modern FCC regulation (if he says "type acceptance", find another guy because that is obsolete), EMC compatibility, and environmental issues (environment as in waterproof encapsulation and corrosion prevention, not environment as in save the whales). I assume you're going FCC Part 15? Or FCC Part 18? What other FCC Part do you intend to list under? Of course if you're going to sell in China they are not so interested in USA FCC rules as they are in their own rules. Which lab are you working w/ for an FCC acceptable "declaration of conformity"? Did you do an error analysis to figure out the required tolerance for each component and how out of tolerance components affect the overall device performance (and fcc rule conformity?) (like, if C5 is 5% low while R65 is 2% high, will it catch fire or merely interfere with airport radars?) Do you have your production line testing gear set up to verify how the devices meet their tolerances? Begun thinking about ISO9000 paperwork yet?
As for the industrial / mechanical engineers job, you need to figure out who the competition sells to, and how you'll sell it as a better product. Frankly, a millimeter is not terribly accurate for a contact microswitch, of course your device is non-contact, and possibly (possibly!) more reliable, or at least probably has fewer moving parts than a microswitch. Then again, IR beams are non-contact too. On the other hand, a machine vision rig is pretty accurate and expensive and can do stuff an IR beam system never could do, and prices are dropping. So you've got a sort of narrow window to market in, between the $2 microswitch and the $300 webcam plus single board computer, and its gotta be something the IR beam interrupters can't do. That market isn't too big, probably because no off the shelf device exists to work in it, or maybe because no one needs it, but either way, no one currently designs machinery that would need it. So, you need some kind of machine designer whom has designed things who can explain how they could have used your sensor device in a past assembly line or whatever, and could intelligently describe what your data sheets need for him to use your device, what the page in the "mouser inc." catalog needs to look like, etc.
Re:Contradictory? (Score:3, Informative)
I am an RF engineer and it is realistic, BUT extremely difficult. Not, IMO, realistic for a hobbyist, even a sophisticated one.
Note that with proper postprocessing and lots of measurements, the GPS system can achieve accuracy close to this ("close" being a single order of magnitude at around 1 cm), and that's on a global scale. With precisely positioned antennas, you can probably do 1mm accuracy over short ranges.
BUT:
1) You're going to need millimeter-precision manufacturing AND positioning of all antennas. Barely within the range of a sophisticated hobbyist.
2) You're going to need some pretty decent signal processing to resolve wavelength integer ambiguities (the same reason achieving centimeter-accuracy GPS requires 1-2 hours of continuous data from a static reference and lots of processing typically.)
3) Multipath is going to kick your ass accuracy-wise anywhere outside of an RF anechoic chamber. Even in an anechoic chamber, you have to deal with multipath and antenna pattern distortion issues due to the subject whose motion you're trying to capture.
If he were making the same claims for an ultrasonic-based system it would be a lot more believable.