Three Unique Approaches to Voice Picking
Zondervan’s $240,000 voice investment is clearly paying off. In the first 13 days that the system was up and running, accuracy improved by 56.7 percent.
Order picking with voice is undergoing yet another transformation. Since its pioneering days in the 1990s, voice picking consisted of predominantly proprietary hardware and software solutions, such as those by Vocollect, using mobile computers embedded with speaker-dependent speech engines.
Then in the early 2000s, vendors such as Voxware started moving away from proprietary hardware and shifted to more open architecture solutions that they embedded in commercial, off-the-shelf mobile computing devices such as those marketed by Motorola and LXE. This open hardware era saw an increase in speaker independent technologies and the rise in multimodal functionality allowing devices to capture data multiple ways, whether via voice, scanning, or RFID.
But over the past three years, the proliferation of high-performance wireless networks and Voice over Internet Protocol (VoIP) phone systems has ushered in what could be a new era in voice. Coca-Cola Enterprises, partnering with Cisco (a leading provider of wireless networks) and Datria (a Lockheed Martin spin-off specializing in packaged voice-enabled enterprise mobility applications) helped to innovate this network-based approach.
“It’s the era of intelligent networks,” says Daniel Hong, lead analyst of customer interaction practice with research firm Ovum and author of the The Guide to Voice Solutions in Warehouse Environments (February 2009). “Voice goes through a physical router that is also connected to a back-end server where all the speech recognition and intelligence resides,” he explains…..continued (click here to download PDF).