Cars—the new home for VPAs

Virtual personal assistants (VPAs) are gaining in popularity as the automotive industry embraces them to connect the car and home to the IoT

Speech recognition is literally a technology created to help machines listen to humans. In its infancy, speech recognition was widely considered a failure, as recognition rates were low and error rates were high. It was common to require users to learn a special vernacular of terms and commands in order to get speech recognition to work. Many automakers tried to structure their speech recognition command dictionaries to be more “natural” but they too often fell short because “natural” speech syntax for one person might be completely foreign to another, even if they both speak the same language. Despite drastically improving during the last few years in consumer electronics, speech recognition as a functionality still lags somewhat when it comes to automotive applications. That said, capabilities here are rapidly improving in the software and hardware realms to closer match consumer expectations. Around the time when the automotive industry started getting good at designing and implementing “more natural” speech recognition solutions, along came Apple and Google with Siri and Google Voice, respectively. While not fundamentally different – user speaks commands, device listens and responds accordingly – the Apple and Google solutions soon proved to be far superior to in-vehicle speech recognition platforms.