Scope of a patent: Electric Imp's BlinkUp

[gcardinal]

Info and introduction to BlinkUp

I have for a while developed a similar solution to BlinkUp. In simple terms monitor on iPhone blinks, device reads data, converts it to WiFi config and connects device to WiFi. And BlinkUp is basically the same thing, but a bit more advance and with its own app and all that stuff. And its patent pending.

So my question is, do their have patent on all solutions where WiFi configuration is transmitted over visible light communication or is it patent on protocol they have developed?

I did send email asking ElectricImp about this, but at the same time, I would like to hear what you guys think about this kind of patent and how big scope can it have?

[Dago]

Info and introduction to BlinkUp

I have for a while developed a similar solution to BlinkUp. In simple terms monitor on iPhone blinks, device reads data, converts it to WiFi config and connects device to WiFi. And BlinkUp is basically the same thing, but a bit more advance and with its own app and all that stuff. And its patent pending.

So my question is, do their have patent on all solutions where WiFi configuration is transmitted over visible light communication or is it patent on protocol they have developed?

I did send email asking ElectricImp about this, but at the same time, I would like to hear what you guys think about this kind of patent and how big scope can it have?

Well naturally it depends on what the patent application is for... Patents are usually very specific and it depends on the actual wording and specifics on what they have patented.

[John Coloccia]

I just spent about 20 minutes out at the website, and for the life of me I can't figure out just exactly WTF this Electric Imp thing does and why it's useful.

[gcardinal]

...
Well naturally it depends on what the patent application is for... Patents are usually very specific and it depends on the actual wording and specifics on what they have patented.

Yes understand that, but at the same time - visible light communication - is nothing new. After all its a just a signal lamp - I just cant understand which part of that it is possible to patent. Source closed app or proprietary hardware design part of light receiving unit - which again is quite basic stuff...

And at the same time, (from what I understand, please correct me if Im wrong) patent pending means that wording is not available to the public...

But when its "patent pending" dose it mean that I cant use this "technology" during patent pending process - or can I start using it and just hope it don't violates patent when they do get patent?..

I just spent about 20 minutes out at the website, and for the life of me I can't figure out just exactly WTF this Electric Imp thing does and why it's useful.

Its kind of a "Nest"-kind of deal, the one Google gave loads of cash for... It makes it "easy"...

[calexanian]

Unfortunately since its not published yet it will be hard to get any information. I just talked to them at maker faire and they seemed to be good guys. I would simply call and ask, under an assumed company name perhaps! Hahahahaha... If you have a system you are developing you may want to join them rather than try to beat them.

[zapta]

Interesting product though the challenge will be getting third party adopting their technology.  They don't have a patent yet, just application so nobody knows that it will cover. The claims can go in different directions from system wide to specific aspects of using camera + screen for serial optical communication (vs QR style parallel data), communication API over standard SD card interface and so on.  It can be as wide as the prior art will allow them and as good their lawyer is (it makes a big difference).

Companies typically avoid discussing patents because of legal implications. If you bring a prior art to their attention they must disclose it to the patent office and if they infringe the penalty is higher. Also, if they explain the scope of the patent to you it may limit the scope of the patent so if they are smart, I expect them not to answer your patent queries and not comment in this thread.

[John Coloccia]

This just seems like so much more trouble than just a simple Bluetooth connection.  I still can't figure out exactly what technology they're selling.  The "blink-up" thing is kind of a yawn.  What's the whole network connectivity thing.  Isn't that just turning the phone into an access point?

[Richard Crowley]

Sending data by flashing light on a screen is not new. There has been "prior art" (as the patent lawyers refer to it) going back at least a decade.
But "prior art" doesn't seem to play much of a factor any more in granting patents. I seriously question whether the USPTO is up to the task any more.
Patent applications are typically as wide as they think they can get away with. Then they break it down into narrower and narrower "claims" so that if they are challenged, they have several "fall-back" positions to fight from.  I think we would be better served if at least half the people who went into law had chosen a different occupation.

[Richard Crowley]

This just seems like so much more trouble than just a simple Bluetooth connection.  I still can't figure out exactly what technology they're selling.  The "blink-up" thing is kind of a yawn.  What's the whole network connectivity thing.  Isn't that just turning the phone into an access point?

I think you missed the main point here.  The brilliance of this scheme is that you can use a very localized (i.e. secure) non-RF communication link to CONFIGURE the RF link. This is especially valuable in small form-factor gadgets that have minimal (or NO) user interface. 

More and more wireless microphone kits (transmitter-receiver) are using this scheme to coordinate setup between the transmitter and receiver.  Because they are now microcontroller-operated, receivers have built-in scanning functions to select the most favorable channel on which to operate. Then, by holding the devices face-to-face, the receiver sends the channel configuration to the transmitter via infra-red light.  See Dave's teardown of his Sennheiser wireless mic.

[gcardinal]

Got email back from Electric Imp guys, here is the patent application:
https://www.google.com/patents/US20140025795

Even has a nice diagram illustrating it all:


It was fascinating to read this patent application - its basic and genius. From looks of it, I think patent will be granted and I will try to find another way of doing it

However I must say, that I find it a bit of a shock, that my humble noob project can`t use something as simple as that... 

[Richard Crowley]

The only reason that a patent would be granted for that is the sheer incompetence of the USPTO.
There is quite obvious "prior art" that should prevent even applying for a patent on this concept.
It is not clear whether the people at the USPTO really are that ignorant, or whether they just don't care.

[NiHaoMike]

What about put a microphone into the device and send the data with (ultra)sound? It even has prior art dating back 3 decades or so.

[PlainName]

There is quite obvious "prior art"

Not sure if it's applicable in the US, but here in Blighty the BBC did this 20+ years ago when they ran the series of computer programmes based around the BBC micro. A small rectangle of the TV screen was used to transmit data to a pickup placed over the rectangle - it is as if this patent is lifted exactly from that but with the TV replaced by a phone.

I guess that if Apple can circumvent prior art by adding 'on a phone' to whever non-invention then this could fly.

Also, it doesn't really matter if prior art does invalidate this: if you're a small developer, where are you going to get the money to fight your case in court? So long as they get the patent, they can go after as many small fish as they figure can keep their coffers topped up, and stiff them for a little less than it would take to fight in court.

[calexanian]

They gave me an Electronic Imp dev kit at Maker Faire. I am going to start playing with it.... When I remember where I put it... Its a small little bugger.. I remember a watch some years ago you programed by holding up to an area on your monitor and it flashed out a code. The idea is not new.

[mjkuwp]

If your project is really little then I don't think electric imp company would bother you.  It is also possible the scope of their patent is narrow.  On the other hand, don't count on the patent office to reject patents that are not awesome.  They approve pretty mundane stuff all the time.


regarding...

I just spent about 20 minutes out at the website, and for the life of me I can't figure out just exactly WTF this Electric Imp thing does and why it's useful.

I have been using electric imp for about 1.5 years, mainly for education and hobby but I have also done one machine for work.

The main benefit is that I get a split-screen IDE where the right side is code for hardware and the left side is code for a virtual machine that runs in the cloud dedicated to that bit of hardware.  It is easy to pass data between the device and the VM (which electric imp calls "agent").  The third component to the solution is a page of javascript/jquery/html and poof- I have a web interface to my bit of hardware.  The data structures in the imp scripting language pass easily to JSON and the language feels a bit like javascript to me so it is easy to move between them and be productive.

electric imp got off a bit on the wrong foot with the old version of their IDE and the first SD-card form factor imp.  However, the new imp002 module and the new IDE are really really good. 

[drasko]

Hi all,
what's the conclusion of this discussion - can the WiFi configuration transmission over the light be used in the projects?

Because if not, someone should tell to Philips: http://www.theinquirer.net/inquirer/news/2353102/philips-shows-off-ethernet-powered-led-connected-lighting-for-offices-of-the-future .

BR,
Drasko

[PlainName]

someone should tell to Philips

Philips aren't sucking power out of WiFi - they are connecting the lights to an Ethernet network via cable. The wireless aspect comes in broadcasting data via the lights so optically enabled devices (like phones) can see Ethernet data without being connected to anything.

Start of the fifth paragraph of the article you link to (my emphasis):

The Ethernet cable-connected luminaire means that Philips...

[EEVblog]

Sending data by flashing light on a screen is not new. There has been "prior art" (as the patent lawyers refer to it) going back at least a decade.

People did this in the 1970's and 80's with CRT monitors.

[Mechanical Menace]

The only reason that a patent would be granted for that is the sheer incompetence of the USPTO.
There is quite obvious "prior art" that should prevent even applying for a patent on this concept.

I saw similar projects at school so agree about the prior art. At least it's not the vague sort of patent that has been commonly accepted since the late 80's. And I still don't understand how software patents can be allowed at all, but especially "black box" style ones.

It is not clear whether the people at the USPTO really are that ignorant, or whether they just don't care.

Through my understanding there is a bipartisan effort to keep the USPTO undermanned. Some of the biggest complainers I've heard about their automatic acceptance of many patents are from the patent office itself.

To be fair it's also not their responsibility to check up on prior art, only prior patents. Prior art may make a patent void but doesn't stop it been issued.

[drasko]

Cool. If you use comparator (or different electrical components) I do not think that this changes subject of the patent and your product is probably illegal (i.e. they pattented transfer of WiFi configuration over the light code - which is weird and probably scandalous patent grant).

The only optimistic thing I think that matters is this (tsken from here: https://www.google.com/patents/US20140025795):
"15. The method of claim 13, wherein said communications module further comprises a standard solid state memory card physical interface with physically compatible electrical connections;
wherein said communications module also comprises an electronic and logical interface that, in at least one mode of operation, is capable of connecting to a standard memory card reader and storing and retrieving memory data according to standard memory card protocols;"

Per my understanding, if you do not store ransfered WiFi configuration on an SD you're good to go.

For the reference: http://assignment.uspto.gov/#/assignment?id=31641-825&q=patNum%3A(8984107)

BR,
Drasko

[aargee]

Isn't this what Timex was doing in the '80s?  ...

http://en.wikipedia.org/wiki/Timex_Datalink

[helius]

But when its "patent pending" dose it mean that I cant use this "technology" during patent pending process - or can I start using it and just hope it don't violates patent when they do get patent?..

While a patent is pending it does not protect an invention. It only protects the invention in case it is approved (and during the process the application may need to be amended, or abandoned and replaced with a different application with a narrower scope). The reason that an inventor prints a "patent pending" notice is that will discourage competitors from copying the design in case the patent is granted. The statutory damages in an infringement suit are naturally going to be higher if the defendant was aware that a patent was pending, and developed a copy, and continued to sell the copy when the patent entered validity. If the defendant knew the patent was pending, developed a copy, but withdrew their product from the market the minute the patent was granted, they have an affirmative defense against infringement. However, this means that the company must write off all of the NRE they spent. So "patent pending" is a real deterrent to copying the product.

[drasko]

For the reference,
one more interesting patent on the subject:
http://www.econais.com/econais-launches-probme-secure-configuration-software/
https://www.google.be/patents/US20130272164?dq=inassignee:%22Econais+Ae%22&hl=en&sa=X&ei=nRQ6VZfIDoe-PKrOgMAM&ved=0CCgQ6AEwAQ

BR,
Drasko

[lapm]

I wonder if optocoupler would qualify as priorart, or optical light barriers that were popular on alarm system decades back...

How ever that docent prevent them patenting protocol they have developed themselfs..

[eas]

The only reason that a patent would be granted for that is the sheer incompetence of the USPTO.
There is quite obvious "prior art" that should prevent even applying for a patent on this concept.
It is not clear whether the people at the USPTO really are that ignorant, or whether they just don't care.

It seems a little premature to pass judgement on this patent without reading it. However, without reading it, I think it is safe to assume that they aren't claiming a patent on using a flashing display to transmit data to another device. Even if they are, it would almost certainly be a broad claim accompanied by many increasingly narrower claims.

As for those narrower claims, some of them are likely variants on the specific case of using a flashing display to communicate WiFi configuration information to a device. That might seem quite obvious, but apparently it isn't, given that it isn't exactly a new or niche problem, and yet all the independent WiFi devices I've used in the past 5+ years inflict some amount of manual tedium in their initial configuration.

[ruffy91]

There is also an IEEE standard for VLC (visoble lights communication). It's used in traffic signaling etc. IEEE 802.15.7

[donotdespisethesnake]

Yes, completely obvious and shouldn't be patented.

The USPTO will approve patents for just about anything these days, no matter how obvious or non-inventive.

The US patent scheme is totally broken. It is basically like the old Royal charters, you pay some money to the authorities and they give you exclusive sales rights, backed by the full legal power of the state. The US government doesn't care much, as long as one US company gets the sales rights, it implements protectionism under the pretense of "IP rights".

[Zero999]

Is it possible to submit an appeal to the patent office to stop a certain patent from being granted? You should make them aware that this is nothing new and has all been done before.

[Mechanical Menace]

Is it possible to submit an appeal to the patent office to stop a certain patent from being granted? You should make them aware that this is nothing new and has all been done before.

Unfortunately not. Your only option after a patent is issued is to take the holders to court. Before it's issued you can inform those applying for the patent that it's prior art and technically they should inform the patent office, but if they don't your only option is to take the holders to court...

EDIT: Though there is a good chance the patent is more meant to protect them from patent trolls than to stop anyone else using the idea. It's become quite a popular reason for people to attain pointless software patents.

[madires]

There is quite obvious "prior art"

Not sure if it's applicable in the US, but here in Blighty the BBC did this 20+ years ago when they ran the series of computer programmes based around the BBC micro. A small rectangle of the TV screen was used to transmit data to a pickup placed over the rectangle - it is as if this patent is lifted exactly from that but with the TV replaced by a phone.

Teletext/Videotext is based on the same principle. A quick lookup shows that BBC did the first tests in 1973 (called Ceefax). I also remember the German TV program WDR Computerclub which used a blink code to transmit computer programs (videodat). Must have been in the late 80s. And my token generator for online banking uses a blink code too (chipTAN).

[zapta]

Is it possible to submit an appeal to the patent office to stop a certain patent from being granted? You should make them aware that this is nothing new and has all been done before.

http://teas.uspto.gov/ccr/lop

[Mechanical Menace]

Is it possible to submit an appeal to the patent office to stop a certain patent from being granted? You should make them aware that this is nothing new and has all been done before.

http://teas.uspto.gov/ccr/lop

I stand pleasantly corrected.

[zapta]

It seems a little premature to pass judgement on this patent without reading it. However, without reading it, I think it is safe to assume that they aren't claiming a patent on using a flashing display to transmit data to another device. Even if they are, it would almost certainly be a broad claim accompanied by many increasingly narrower claims.

+1

The patent protection is given based on the claims which are typically a list of AND conditions (the more, the narrower). To fit a claim all the conditions must met, so if one of them is, let's say, is about IR communication it doesn't mean that the patent protects IR communication in general.  To evaluate if a patent is valid one must look at the totality of each claim, not just one component.

It's ok for a patent to be based on multiple existing concepts that are combined in a non obvious way (non obvious at the time of the filing, not now in retrospect). I think that the magic number is 3 (combining two existing concepts is insufficient) but I am not sure, not a lawyer.

BTW, the Amp Hour episode with the guy from Electric Imp touches on this patent briefly.

[LabSpokane]

EDIT: Though there is a good chance the patent is more meant to protect them from patent trolls than to stop anyone else using the idea. It's become quite a popular reason for people to attain pointless software patents.

Bingo. It's cheaper to buy 50 "patents" in the US than to defend one patent lawsuit.

Is our patent system utterly broken? Yes it is.

Don't read ElectricImp's patents as anything other than defensive. It's what companies must do in the US to have a chance of being attractive to investors/buyers.

[PlainName]

EDIT: Though there is a good chance the patent is more meant to protect them from patent trolls than to stop anyone else using the idea. It's become quite a popular reason for people to attain pointless software patents.

Bingo. It's cheaper to buy 50 "patents" in the US than to defend one patent lawsuit.

There is also a third reason: it is better marketing to be able to say your product is patented (implying it is one of a kind), although 'patent pending' can sometimes sound better to punters.

Indeed, sometimes it is worth getting a patent for something which the product doesn't actually rely upon but can be associated with the product. The roughly $5-10,000 cost could easily be swallowed by a single advertising blitz but lasts 20 years...

[Maxlor]

This is an old thread, but a colleague and I have been discussing this patent today, and I saw that there's no definitive answer to "can I do blinkup in my own project" in this thread. Therefore I'd like to point out a couple of things about patents:

- Only the section labelled CLAIMS matters. Everything else, the pics, the examples and explanatory texts are irrelevant and with non-US patents can sometimes even be deliberately misleading or confusing.
- For some device to be covered by a claim, it must have all the attributes mentioned in the claim. This means that very short claims are a lot more valuable (and harder to circumvent) than long claims. For example: a patent covering "a vehicle riding on wheels" would cover all cars, whereas one that covered "a vehicle riding on wheels with a combustion engine having 9 cylinders" would cover pretty much none of them.
- Prior art has to also satisfy all aspects of a claim, not just some of them.
- Assume the main claims (printed black not gray by google) are given. The other claims (the ones starting with "the method of claim 1, and...") are there in case the patent office doesn't deem claim 1 on its own patentable. These additional claims further restrict claim 1, making the patent narrower, and thus less valuable. It may happen that the patent office only grants claim 8 for example. In this case, this hasn't happened, the independent claims 1 and 13 were granted, the others are irrelevant.

So, what is Electric Imp patenting? In my own words:

a device with:
- circuitry that runs software, AND
- is able to store a unique ID, AND
- is able to transmit that unique ID, AND
- has both a radio and an optical interface, AND
- has solid state memory for both code and configuration data, AND
- uses the optical interface to load network configuration data, AND
- connects to the network specified in the configuration data, AND
- loads code based on the unique ID or a token, AND
- sends the unique ID or a token to a server, which uses that ID/token to select code (note, it doesn't say where this code runs), AND
- the device communicates over the network.

If you build your own device, and want to make sure that it doesn't fall under this patent, it is sufficient to make a single point in that list unapplicable, like not having a server pick different code, or not transmitting a unique ID.

So in a nutshell: can you copy the blinkup process? Yes. Can have a server run individual virtual machines for each of your devices? Also yes. Can you combine blinkup and server virtual machines? No.