Voice recognition: Is it now mature enough?
Filed in archive Technology on December 28, 2008
Silicon Republic has recently published an article on new advancements in WMS solutions and one of them was voice recognition technologies. This technology which has been introduced in the last five years, claims to improve the lead-time and cost of picking, material handling and pallet management process in the warehouse.
The companies promoting the technology seem to have thought about the adoption process as well, as an implementing consultant company explains:
"The speaker is tied specifically to the voice of the user. The employee is given 100 items to pick, for example, and told where they are. When he picks an item up he says the product code and the computer confirms if it's the right one.
"As employees move around they quickly get used to the system and get a rhythm going. They find it much easier than dealing with paper orders as their hands and eyes are free all the time. Order pickers we've spoken to say they wouldn't go back to the old way of doing things."
However, I have seen this technology in action and even though it seems very cool, I think it still needs some time to get more mature: As an example in the environment with a lot of noise, the devices sometimes fail to deliver the promised quality.
What do you think?
Permalink: Voice recognition: Is it now mature enough?
Tags: scm supply chain management wms warehouse reliability technology 2008 supply+chain
Vote for Voice recognition: Is it now mature enough?:
|
Rating: 9.20 out of 5 vote(s) cast.
|
Response from:
Aitor Urzelai
(12/28/08 12:13pm)
I agree with you. In Spain, pick to light is quite well implanted, but pick to voice is in a much more premature stage. Very few enterprises are using it. Some others are testing it, but in general, technical problems in voice recognition are still quite common.
Response from:
Doug
(12/31/08 12:36pm)
You seem to be discussing "mis-recognition" problems. Technology is not the only source of mis-recognition. For example, the worker may not have the headset microphone in the right place. The worker may be yelling at the system instead of talking in a normal voice. The headset may be only average in its noise canceling capabilities. Additionally, there are three distinct types of speech recognition technology in today's "voice picking" market - all viable but different. Two are thick client solutions: (1) speaker dependent where each user creates their own voice templates, and (2) speaker-independent which works out of the box. There is also a thin-client technology which is very common in contact centers: (3) network-based speaker-independent. All of these speech technologies have thousands of users successfully deployed in noisy environments. Good dialogue - but not quite as simple as it may originally sound.
| RSS |  |
 | |
| Yahoo! |
|
| Addthis |
|
| Bloglines |
|
| Follow us on Twitter! |
Most Popular
Basics
Best of
Best practice
Book Review
Did you know
Education
Employment
General
Green supply chain
Green supply chains
Guest Column
Implementation
Interviews
Market Overview
merger and acquisition
Misc
News
Partnerships
Point of view
Practical Tips