SEARCH searchBY TOPIC
right_division Green SCM Distribution
Bookmark us
sitemap
SCDigest Logo
distribution

Focus: Distribution/Materials Handling

Feature Article from Our Distribution and Materials Handling Subject Area - See All

From SCDigest's On-Target E-Magazine

- March. 11, 2014 -

 
Supply Chain News: Understanding the Software Components of a Voice System in Distribution


Many Pieces May Be Deployed, Depending on Needs and the Capabilities of Existing Systems

 

 SCDigest Editorial Staff


The following article is an excerpt from Supply Chain Digest's new White Paper entitled A 360-Degree View of Voice Technology in Logistics 2014. That report can be downloaded here.

There are actually quite a number of different pieces of software that are or can be needed to make Voice work. Exactly what pieces a given company needs for its Voice deployment varies of course with application requirements, existing software, the supplier a company decides to go with, and more.

SCDigest Says:
start
A new type of software involves converting existing RF data fields from the WMS into Voice commands, with the reverse being true for users' Voice responses.
close
What Do You Say?
Click Here to Send Us Your Comments
feedback
Click Here to Post or See Reader Feedback

Below, we summarize the different Voice-related software that are common in distribution center applications.

"Host" System: In almost all cases, there is one or more "systems of record" to which the Voice system is connected. This is most frequently a Warehouse Management System (WMS), but can be an ERP/Order Management system or something else, like independent vendor compliance/audit software or Lab Information Systems.

Increasingly today, the WMS solutions of leading best of breed and ERP providers have "out of the box" support for some but not all Voice solutions. However, few have complete Voice support across all of their existing radio frequency applications. Most picking applications (full case, split case, etc.), for example, would be supported, but receiving or truck loading processes may not. That means companies wanting to Voice-enable those other areas either need to have the WMS provider do some additional work or use the application management software from a Voice software provider (see below).

Voice Recognition Software: The heart of a Voice system is naturally enough the software that actually recognizes the words spoken by a user and translates that into data for processing by the application programs, as well as translating host software data back into audio responses.

This is an area where the software has made significant progress over the past decade, greatly improving recognition accuracy and speed. That in turn reduces implementation times and later operational problems, helping to further decrease the time-to-value and increase the ROI from Voice investments. This is especially true in dealing with "noisy" DC environments.

In addition to the environmental noise factors, consider the challenges of accurately recognizing in real time what words perhaps hundreds of workers in a distribution center are saying, each using different tones, accents, and more – and the variability usually seen even in how a given worker may pronounce each word at a given moment.

While today all Voice software vendors have solid recognition capabilities, there are differences among them, especially in how they might perform in the specific environmental conditions of a given distribution center.

Here the debate is also joined over "speaker-dependent" versus "speaker-independent" Voice systems. In the latter approach, each Voice system user "trains" the system on the way he or she pronounces the 100-200 words typically used in a given Voice implementation. Speaker-independent systems use phonetic recognition to identify words regardless of speaker, such as is required in call center applications where tens of thousands of different customers may be dialing in.

There are pros and cons of both approaches, though nearly everyone agrees in recent years tremendous progress has been made in speaker-independent technology. The major downside of the speaker-dependent systems is the time required (20-30 minutes) for each user to train the system on their individual pronunciations, though some argue that comes at the benefit of greater accuracy/fewer failed recognitions.

Potential Voice users simply need to be aware that there are different approaches and weigh the pros and cons along with all the other factors involved in selecting a Voice hardware and software provider.


(Distribution/Materials Handling Story Continues Below )

CATEGORY SPONSOR: LONGBOW ADVANTAGE - JDA SUPPLY CHAIN CONSULTANTS

Download Longbow Advantage

Business Briefs

 

 

The Keys to WMS Success,

Maximizing JDA WMS

Performance and More

 

 

 

 
 

Voice Application Management: Many distribution centers run older, perhaps even legacy WMS systems that do not have any native Voice support. Or, as mentioned above, an installed WMS may have Voice support for order picking, but not for other application areas.

If that is the case, potential Voice users have four main choices:

1. If possible, upgrade the existing WMS to a version that better supports Voice;

2. Build true Voice support into the existing WMS (not common);

3. Use floor-level applications from a Voice vendor;

4. Use one of the new generation "screen mapping tools" to achieve connectivity (see below).

Most Voice software vendors have applications for areas like order picking, or can create applications for other areas. These applications contain business logic, workflow configuration and more to enable a specific process. At a high level, a common scenario would be for the WMS to send a block of work to the Voice application, such as a group of picks by item, quantity and location.

The Voice application would take that data and issue pick commands to workers, confirming the work done, etc., eventually sending that information back to the WMS. This solution can work quite well, and has been deployed in probably thousands of distribution centers, but has the downside of generally not being truly "real-time" in communication with the WMS.

But for users that have limited Voice support in their WMS, deploying a robust Voice application that can optimize workflows and even provide real-time operational metrics is in fact often the smart choice.

Voice Screen Mapping: In the past two years, a new type of software to achieve integration of Voice with WMS's lacking Voice support has been brought to market. This involves converting existing RF data fields from the WMS into Voice commands, with the reverse being true for users' Voice responses.

This is generally achieved using what might be called a "screen mapping" approach, and in fact these solutions can often be used to actually simplify the existing prompts and screen flow that exists in the WMS RF application. Many existing RF programs have more screens then they need, for example, and can be collapsed into a brief series of Voice commands and responses.

This technology can also be used to achieve very rapid and relatively inexpensive WMS connectivity.

Voice Management Consoles: Most Voice software will have control center software that allows customers to see and manage their Voice users and equipment, update user configurations, measure performance and other capabilities.

When dealing with Voice vendors, it is sometimes difficult to identify all the software elements that will be required for the total solution, and how those applications are prices. Potential users should insist on full disclosure here.

Again, to download the full white paper, go to A 360-Degree View of Voice Technology in Logistics 2014


Anything to add about Voice-realted software? Let us know your thoughts at the Feedback button (email) or section (web form) below.

Recent Feedback

 

No Feedback on this article yet

 
 
.