The Talking Warehouse: Understanding How Voice Unleashes Higher Performance in Product Distribution

An eBook by: Roger G. Byford and David Maloney

Chapter 2: Demystifying Voice in the Warehouse

Let’s be honest, this “human speaking to computer” type of experience has not always been positive. For instance, we may have phoned a customer service center and been asked by the electronic “receptionist” to state our reason for calling from a list of options. We spoke a response, and often the system failed to fully recognize what we were saying. We have to repeat the process, practically shouting at the phone and over-emphasizing our diction until the system finally understands. It leaves many of us frustrated and shaking our heads at such technology.

It is little wonder, then, that many supply chain managers fear that voice-directed workflow systems will offer similar results. If their experience with voice has been tainted by their encounters with consumer speech recognition systems, then they probably have low expectations for how well voice can succeed for their distribution operations in a “human talking to computer” type of arrangement. Little do they know that not all speech recognition systems are created equal.

In this chapter we will demystify voice and highlight its unique design to generate productivity and accuracy improvements in a distribution environment. We will pull back the curtain to reveal how there are different types of speech recognition systems that vary widely in their abilities to process information and provide desired accuracy. We will examine how some speech recognition systems have very high recognition rates in comparison to others, and how each is built to recognize various sizes of vocabularies, numbers of speakers, and types of environment (e.g., noisy versus quiet). We will look at how some speech recognition systems are built on sounds within words, while others use words themselves as the building blocks.

We will also explore why the speech recognition technology used in warehouse-based voice systems is a completely different animal from what we have experienced in the consumer arena. In particular, we will investigate the techniques the builders of warehouse voice systems use to create recognizers with very high accuracy.

The environment in which voice systems operate can greatly affect their performance. We will see how companies have engineered their products to match the demands of industrial warehouses. In this chapter we will also discuss how hardware, software and computer platforms can affect how well a speech recognition program functions.

Finally, since this book is directed to supply chain professionals, we will look at how the various pieces of the voice puzzle work together within the distribution environment to provide a system that is highly accurate in its ability to recognize a user’s speech, using the attributes of speech interaction to build a productive workflow – thus driving increased business results.

This chapter focuses on how a computer understands speech from a human. To carry on a conversation, the computer obviously has to be able to speak to the human as well. We’ll review computer speech output in Chapter 6 of The Talking Warehouse.

>>Download The Full Chapter By Filling Out The Form Above