Patently Apple, as is typical for a site whose mission is "celebrating Apple's Spirit of Invention," mainly rehashes what Apples says about its inventions in either the final patent or, as in this case, its patent application. There's rarely an attempt to provide additional context or third-party explanations or assessments about the technologies mentioned by Apple.
What Patently Apple's post describes is a system of audio sensors hidden in the screens or surfaces of a wide range of devices. Essentially, the system appears to be a way to detect sound waves and then, you know, do something with them -- they'll act as "inputs" to the computer or possibly be used to create some kind of "output."
Or as Patently Apple puts it: "One or more sensors are coupled to the display and configured to detect vibrations of the screen generated by sound waves impacting the exterior surface of the screen. A processor in communication with the one or more sensors is configured to generate an output representative of sound waves."
What does that actually mean? As is often the case with Patently Apple posts, one has the sneaking suspicion that the writer doesn't actually know, and fakes it by paraphrasing or quoting from Apple's document. For example: "In a third embodiment [apparently a fancy word for 'example'], Apple's invention may include a method of operating a computing device. The method includes obtaining an electrical signal corresponding to vibration of a screen of the computing device resulting from sound waves impacting the screen and filtering the signal to remove noise components. The method also includes generating an output signal representative of the sound waves that impacted the screen."
We're just guessing here, but this seems to mean that you speak, the screen vibrates, and the computer "translates" the vibration into some command like "call Mom" or "buy two tickets for the 4 p.m. show of 'The Hobbit.'" But Apple's Siri voice assistant already does most of that, more or less effectively, with no need for audio sensors and all that other stuff. Nobody confuses Siri with sonar. And there's nothing in Patently Apple's post to explain why the integrated audio sensors and other stuff would make voice control more accurate than what Siri offers today (and in the future).
If you want some idea of what sonar actually is, you can check out the Wikipedia entry. Active sonar sends out a sound and listens to the echoes, and then uses calculations to figure out the location, size and other features of an underwater object. Passive sonar doesn't do any pinging: It's just listening for sounds, such as those made by a submarine, and figures out what the sounds represent, like, a submarine.