As individuals around the world become more accustomed to using mobile computing devices in hands-busy, eyes-busy environments, speech recognition technologies within these devices is expected to gain considerable traction within the next five years.
A new report, The Proliferation of Innovative Speech Applications in Mobile Computing, predicts the global market for advanced speech recognition (ASR) in mobile handsets will increase from $32.7 million in 2009 to $99.6 million in 2014. That for ASR in in-vehicle telematics is expected to grow at a similar rate, from $64.3 million in 2009 to 208.2 million by 2014.
Speech applications in mobile computing typically enable the user to control a device vocally —a feature known as command and control. Other common applications include voice-dialling and voice-search, in which a search engine can be controlled by voice, and voice-input, in which the user verbally inputs data into the device.
Regulations spur uptake of speech apps
Hands-free laws are gaining traction around the world, restricting the use of mobile devices while operating vehicles. Countries as diverse as Australia, Britain, Chile and the Philippines, among others, have enacted legislation prohibiting heavy interactions of mobile handsets while driving. In the US, 15 states restrict handset use when driving, allowing for the use of hands-free systems. Vendors have been pushing command and control and SMS transcription applications to alleviate these issues.
Providers of in-vehicle telematics, which are entertainment or navigation systems factory-installed in automobiles, and PNDs use a combination of ASR and text-to-speech (TTS) technologies. Common functions include address input, music playlist management, and address read-out.
Personal navigation devices (PND) has been a strong market for speech applications. The expected reduction of speech applications in PNDs is symptomatic of the flagging popularity of the device due to a number of factors, including reduced consumer spending and new cars having GPS built-in.
New ways to deliver speech in mobile devices
Currently, most speech applications use embedded ASR – that is, speech recognition integrated directly into the device itself. However, there has been an increased number of applications that use networked-based ASR to improve recognition rates and to speech-enable internet searches. Datamonitor notes that the advent of app stores, such as the one associated with Apple’s iPhone, offers a viable channel-to-market for smaller providers of speech applications and increased visibility for innovative speech applications.
In future mobile apps will use a hybrid of embedded and network speech. Vendors are offering speech application as suites instead of as point applications. To harness the capabilities of mobile devices, we’ll see solutions that combine embedded speech for features such as command and control and network speech that enables internet search.
Not just for grey collar jobs
Mobile speech applications used by businesses are typically relegated to gray collar jobs that require physical activity. Datamonitor’s findings suggest that because enterprises typically don’t prioritize mobilizing their white collar workforce, the market for speech applications in this sector is low.
Conversely, there’s considerably more upside for speech applications targeted towards mobile field workers and warehouse workers, who consistently operate in a hands-busy, eyes-busy environment, and for whom speech applications can speed the completion of field force or warehouse-related duties.
For example, warehouse workers constantly move and lift inventory. Having to log inventory via a manual process decreases productivity and distracts the worker, which increases the chance of accidents. Warehouses that invest in “voice picking” applications, in which inventory is logged by voice, find that workers can accomplish many of their tasks without having to break concentration from his lifting duties.