Imagine checking the charge status of your electric car while watching TV simply by making a voice inquiry towards the TV. The e-car in the garage or parking lot of your apartment building responds via the TV that it is fully charged. The TV "hearing" this exchange assumes that you want to go out asks if you want to record the show that is on. You answer "yes" and head out the door. As you leave the lights dim behind you, the heating is turned down, the TV turns off and the door locks behind you.
Everyone expects phones to interoperate but supporting voice connections between other types of electronics, such as a TV and home lighting control or an e-car, isn't so obvious. For some time TI has evangelized a vision for voice that says that nearly every connected device will be voice enabled. This is a far reaching vision that we truly believe in. Since the vision is so encompassing it should be obvious that no company can achieve it on its own. There are three key enablers that will help make this vision a reality. First is having a wide choice of embedded processors and SOCs (systems-on-a chip) that are appropriate for the various connected devices that will be voice enabled. Second is the existence of standards that provide a well understood target environment that encourages interoperability across a diverse field of devices. Third is support for open source development across the hardware platforms and key enabling software.
The wide choice of processors mentioned as the first prerequisite is the easy one. SOCs are available for most high volume applications and there are numerous choices ranging from embedded microcontrollers to multicore devices available for emerging applications. Standards and solutions exist for interoperabiltiy at the voice level in the form of codecs and voice activity detectors; but the ability to register a non-phone device, convey its capabilities to other devices in the consumer's network in a secure manner and to interact using voice commands remains as work to be done. OpenRemote.org is one organization that seems to be heading in this direction and it complies with the third requirement for open source development.
If you consider the scenario described above it will be obvious why open source is required. No single company regardless of its size or scope can enable a scenario like this on its own. Cooperation and interoperation is required from the electric utility grid, the e-car manufacturer, the manufacturer of the charging station, the apartment building management, the TV manufacturer, the lighting and access control manufacturer and the HVAC controller at a minimum. It isn't the "free" nature of open source or even the platform independence that it promises that matters here. What matters is the formation of a development community with a shared vision. Only through the combined efforts of a community of developers, each with its own area of expertise can scenarios like this be achieved. No single organization needs to solve the entire problem. The open nature of the community allows the applications to continue to grow and expand in many directions simultaneously.
Obviously this will be possible without using voice but what better, more natural user interface do we have? Interconnecting our voices with our environment in such an immersive way simplifies the interoperability and control challenge for the user and creates many opportunities for our community of innovative developers.