Siri can no longer be considered a startup since it is now part of Apple.  However, not long ago, Siri was one of the most exciting young companies participating in the Apple app store.  By now, everyone is eagerly anticipating Apple’s new Siri personal assistant in the and about to be released iPhone 4s.  It is being said that the Siri voice recognition technology is far ahead of its competition.  So it seems like the competitors have some catching up to do.  After reading about Apple’s history with Siri, I became curious what kind of IP protection has been built around the product.  However, identifying patents relating to Siri proved challenging.

Siri was founded in December 2007 by Dag Kittlaus (CEO), Adam Cheyer (VP Engineering), and Tom Gruber (CTO/VP Design), together with Norman Winarsky from SRI International’s venture group.  It was a spin-out from the SRI International Artificial Intelligence Center, and is an offshoot of the DARPA-funded CALO project, described as perhaps the largest artificial-intelligence project ever launched.  The company was acquired by Apple on April 28, 2010.

Apple’s Siri IP

A search of the USPTO assignment database revealed that there are no patents or patent applications with Siri, Inc listed as an assignor or assignee.  However, an inventor search for the Siri co-founders reveals a number of patents and patent applications.   Of these, only one has been assigned to Apple to date, U.S. Patent Application 2007/0100790 (“the ’790 application”).  The application was assigned from SRI International directly to Apple, Inc on August 25, 2010.  It is interesting to note that Siri, Inc apparently was never assigned this application prior to its acquisition by Apple.

The ’790 application originally claimed a method of “building an automated assistant.”  However, there have been several rounds of back and forth with the USPTO, forcing Apple to narrow the claims.   Apple filed a request for continued examination (“RCE”) in September of 2011, so it is possible that under the new rules for RCEs the Examiner will not pick up this application again for as much as a year.

A PCT International application relating to the Siri technology was recently noted by (If you are interested in the technical implementation of Siri, I recommend taking a look at Unwired’s post).  This International application revealed a second unpublished U.S. Application No. 12/987,982 (“the ’982 application”), which appears to be assigned to Apple, but has not yet been examined.  In the International application Dag Kittlaus, Adam Cheyer and Tom Gruber, among others, are listed as inventors.  It is interesting to note that the PCT application claims priority to the ’982 application, as well as a provisional application filed in January 2010.  Thus, the earliest known application directly related to the Siri technology was filed just 4 months prior to Apple’s acquisition.

Since the ’982 application is unpublished in the U.S., we do not know for sure what the claims are directed to.  However, the International application likely has claims that are very similar to those pending in the U.S. counterpart and claims:

An automated assistant operating on a computing device, the assistant comprising:

an input device, for receiving user input;

a language interpreter component, for interpreting the received user input to derive a representation of user intent;

a dialog flow processor component, for identifying at least one domain, at least one task, and at least one parameter for the task, based at least in part on the derived representation of user intent;

a services orchestration component, for calling at least one service for performing the identified task;

an output processor component, for rendering output based on data received from the at least one called service, and further based at least in part on a current output mode; and

an output device, for outputting the rendered output.

This claim appears to be directed to a very basic, high level, description of the overall Siri personal assistant software.  The claim requires an input device and an output device (the iPhone’s microphone, speaker and display).  Upon further review (Thanks, Ali H), it appears that the language interpreter component is in fact on the iPhone device (and not the text-to-speech component, as I first believed).  According to the specification, the language interpreter component received the speech-to-text output processed by Nuance and determines a parse result.

Other relevant IP

Also of note is that both Adam Cheyer and Norman Winarsky were clearly working on predecessors to what became Siri while at SRI International.  Several patents and patent applications related to  voice input technology with either Cheyer or Winarsky as inventors are currently assigned to SRI.  These patents and applications are listed below.  However, I do not know whether Apple has licenses to any of these assets.

Relevant Patent Applications with Adam Cheyer as inventor:


Relevant Patents with Adam Cheyer as inventor:

1 6,757,718 Mobile navigation of network-based electronic information using spoken input
2 6,523,061 System, method, and article of manufacture for agent-based navigation in a speech-based data navigation system
3 6,513,063 Accessing network-based electronic information through scripted online interfaces using spoken input

Relevant Patents with Norman Winarsky as inventor:

1 7,844,254 Method and apparatus for collaboration and media access using mobile communications devices

Also of note

If you are interested in patents and the iPhone, PatentlyApple has an interesting blog post looking at Apple’s camera patents that have been implemented in the iPhone 4s.