Occasionally, prospective voice users will ask us if we offer a voice over IP (VoIP) version of Jennifer. They are typically asking because they want an option that runs on lower-cost phones rather than industrial terminals. And it’s a fair question, since we do offer Jennifer in a wide variety of architectures and configurations, and on a wide variety of devices, including phones.
But when it comes to whether we offer VoIP, the answer is definitively “No.” And that’s a very purposeful decision on our part. The way our voice systems are architected ensures a better user experience, a simpler and less expensive network infrastructure, and richer functionality than VoIP.
Here’s a little more detail about the differences between Jennifer and VoIP systems:
Voice over IP treats the voice-user interaction as if you are making a phone call and, as such, every voice communication must go back and forth over the RF network as a VoIP transmission. When the user speaks, the system sends the audio over the network to a server that contains the voice recognition software and the application (the user’s assignments). Once the speech recognizer on the server interprets the voice, the application determines what to say to the user, and then it sends another audio file back across the network with the user’s next instruction
Not only can this add seconds to every transaction, but this means that if there is any network trouble or slowdown, the worker will be stopped in her tracks. In order to reduce these stops to production, the RF network in the DC needs to be more than doubled (compared to even a typical scanning application) to handle the huge amount of data (VoIP transmissions) going back and forth. Furthermore, since there’s often a delay in recognition, these types of systems can be very frustrating for the end user.
Additionally, on the IT side you would need technical expertise on maintaining a VoIP telephony environment, which is an often overlooked expense of VoIP warehouse solutions.
In contrast, Jennifer voice solutions use only a traditional wireless network to send data, and the voice recognition engine – Serenade – and the voice application live on the mobile device. This means that even if network coverage drops, the user can continue working. While the coverage is full, the device sends real-time updates back to the host system as picks are completed and, if the coverage drops, those messages are queued and sent back to the host as soon as coverage is adequate. Additionally, because the recognition and Jennifer’s audio commands “live” on the device, the amount of data being transmitted over the wireless network is far less than in a system transmitting audio data, so Jennifer requires much less bandwidth.
Back when the first VoIP systems were put into production, the main benefit to adopters was that they ran on phones instead of industrial terminals, which meant device costs were less. These devices were admittedly less expensive, but also resulted in a loss of functionality – they were voice-only without scanning or large screens. Today, that cost advantage has disappeared with the advent of Jennifer running on Android smartphones, which offer more functionality (scanning, visual display, touchscreen) and fewer risks (buffers against poor RF coverage, recognition and logic on the device itself, etc.).
So, maybe the full answer to the question, “Do you offer a voice over IP solution?” should really be, “No, and you probably don’t want one.”