Microsoft has filed for a patent on a new technique that allow users of mobile devices to silence them by delivering a firm whack. From The Register:

Patent application 20120231838, aka CONTROLLING AUDIO OF A DEVICE, aims to patent:

“A method comprising: in a mobile communications device: receiving information indicative of acceleration of the mobile communications device; determining correlation between the information indicative of acceleration of the mobile communications device and exemplar whack event data; and based at least on the correlation, controlling an audio signal of the mobile communications device.”

Now my HTC Desire HD already uses accelerometers to silence a ringing phone on my desk simply by turning it over.

To me this seems like entirely the same principle, only a different set of data from accelerometers will be used to trigger the silencing of the ringer.

tl;dr - would this count as prior art, or is it not specific enough?

  • 7
    This DIY project constitutes clearer prior art: it's exactly the same principle, all it's missing is that the device is a mobile communications device. This application is an obvious application of this prior art. Sep 18, 2012 at 17:57
  • 2
    @Gilles, can you post that as an answer? It's really additive.
    – Jaydles
    Sep 20, 2012 at 16:17
  • 1
    The HTC Desire operation may not be a "whack event." The DIY project is in the ballpark, but the DIY project is missing feedback to signal a changed state into silent mode. (Silencing the alarm is not sufficient.) It is missing much more than just the device type. See my comment in response to @Jaydles below.
    – user96
    Sep 21, 2012 at 5:14
  • 3
    [Continuation. My editing attempt went on too long.] I am not saying the patent is unassailable, just that the options offered are far short of knocking out the 20 inventions claimed in this patent. Similarity is a threshold for consideration as prior art - not a final answer. The confidence expressed on this page is misplaced.
    – user96
    Sep 21, 2012 at 5:24
  • 1
    This question is kinda what's up for discussion here: are whacking, swirling, rotation, tap patterns, etc. are all similarly trivial, previously demonstrated and well-known? If you replace "whack" with "gaze lovingly at your phone" is that any more novel? How about "think really hard about doing harm to your phone"?
    – Shog9
    Sep 22, 2012 at 17:36

10 Answers 10


I have an app called Shake2MuteCall on my Droid2 for awhile now. I hit my phone and it stops ringing.

  • This seems like a great piece of prior art, especially since the last update of the app happened in May of 2010, a year before MS filed the application Sep 20, 2012 at 16:40
  • 8
    This even touts the ability to "slap your pocket". If that's not prior art I don't know what is. Unless whacking is a different thing to slapping.
    – Jookia
    Sep 20, 2012 at 22:17
  • 2
    The idea that this site might have defeated a M$ patent on prior art grounds is just ... titillating.
    – g33kz0r
    Sep 21, 2012 at 18:02
  • Please see an answer I posted today - the application got a non final rejection in Nov. as anticipated by another patent.
    – George White
    Jan 16, 2013 at 23:07

In context with other examples of prior art (such as Bump's application which allows you to trade contact information by knocking two phones together or the N900's Shake to Control app which controls music with similar movement) this seems like a potentially valuable piece of prior art.

In this case, Microsoft appears to have taken two well known concepts...

  • Using the accelerator to detect motion in a single plane, and

  • Silencing a ringer based on sensing motion),

...and combined them in a very obvious way.

Given the obviousness of this combination and wealth of examples of other devices with very similar features, it seems nearly impossible to deem this patent to be sufficently novel and non-obvious to be approved.

  • 3
    The act of combining two data sets to produce something unique, even in an obvious way does not make it less patentable, eg. using 2 fingers on a touch screen instead of one', is obvious but patentable. Similar features do exist, but the patent is explicit (whack) in that it does not try to overlap those similar prior-art inventions.
    – Neil
    Sep 21, 2012 at 6:54
  • 1
    @Neil Combining two (or even more) known processes or devices in an obvious way does not pass the obviousness test. In order to be patentable, such a combination has to be non-obvious, for example when one of the components is well-known in one field but is new in the field of the invention. Sep 23, 2012 at 1:39

"Whack Gestures" (PDF) published by Hudson et al. at TEI 2010 seems to already cover this claim:

While we were initially motivated by ESM applications, we believe other interactions could also benefit from this approach. Perhaps most compelling of these is quickly responding to (or silencing) a ringing cell phone.


Gestures are performed by firmly striking the device, e.g. with an open palm or heel of the hand moved towards the waist – an action we refer to simply as a whack.

The whole paper is about these whack gestures, so it covers also implementation details (like processing of accelerometer data) and more complex gestures. Therefore, at least a few of the patent application's claims should be covered.

  • 1
    I think your evaluation is on the money, and the "Whack Gestures" paper will almost surely be useful in an examination of the patent claims.
    – user96
    Sep 21, 2012 at 18:54

While it's not silencing a communication device, surely the silencing of an alert noise would even date back to the old alarm clocks which were designed to be thrown at a wall to turn them off. Had a friend had one of those back in the 1990s. Might not be a direct example of prior art, but modifying it to be a communication device rather than an alarm clock would seem a fairly obvious development.



Look at the dependent claims and consider my answer to this question: Prior art for using a camera in self-driving cars.

You cannot do a full prior art analysis just by looking a one independent claim, especially not the first claim of a patent.

In addition, you have to read the entire specification of the patent to determine whether any of the terms in the claims may have uncommon meanings attached to them. The inventor is given latitude to "act as his own lexicographer" and assign peculiar meanings to terms. When such terms are used in a patent claim, it may not be apparent to the casual reader that the claim has a meaning that differs from what one would normally expect.

  • 1
    Whilst you have a great point, how is this relevant to this particular case? Sep 20, 2012 at 15:41
  • 2
    user96, is this a reference to the @alex miller answer? If so, it'll be clearer as a comment. One other note, as your expertise here is really helpful - if you can articulate what is missing from the answer, (i.e., "that's good, but the spec refers to a potato-powered machine, so non-potato examples may only narrow it"), it'll help the responder try to improve or flesh out where possible.
    – Jaydles
    Sep 20, 2012 at 16:20
  • 2
    I did not see anything like an "exemplar whack event" in any of the proposed prior art. It is not clear that a "whack event" refers to just any change communicated by the accelerometer. No one mentioned "feedback" as described in claim 15. "Bump" works by communicating synchronous bumps back to a server. None of the other art claims to operate from data taken from a single axis (claim 9). "It must be that way" and "Everybody knows" do not qualify as prior art without more detail. User-trained, editable, adaptable whack events (claims 2 and 3) are not addressed. QED is not appropriate.
    – user96
    Sep 21, 2012 at 4:59

A non final rejection was issued in November. Besides a rejection due to non patentable subject matter, 19 of the 20 claims were seen as anticipated by us2008/0165022, Hertz. The remaining claim was seen as obvious under Hertz in light of examiner's knowledge. You can look it up in public PAIR.

Also Hertz has been issued as US7671756 in 2010. I did not look at his allowed claims.


There are apps for old nokia N95, such as the ShutUp app by Samir, that do similar things, like putting the phone into silent mode when triggered by an accelerometer action.

  • 5
    If you could cite some specific examples and describe how they relate to the claim in question, this would be a lot more useful. Please consider expanding this answer to include such details!
    – Shog9
    Sep 20, 2012 at 18:02
  • 1
    – astro
    Sep 20, 2012 at 22:32

This effect can be achieved (to a point) using the "Tasker" Android application. While tuning it to work just right may require a dedicated app, this ability has existed for a long time.

  • 5
    Any turing complete programming language can simulate any other turing complete programming language; just because Tasker can be programmed to do it does not make it a prior art.
    – Lie Ryan
    Sep 20, 2012 at 17:40
  • Hi zeel, and welcome to the site! Sep 24, 2012 at 2:42

My alarm clock responds to exactly the same stimulus, and I believe this has been an approach for decades.

  • 2
    Rob G: Thanks for the answer, but this would be difficult to submit as prior art. Would you consider expanding you answer to include such details as what the product is, etc? Thanks. Sep 21, 2012 at 16:10

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .