Voice recognition for WMS

Posted by anthony.micali on 05-Nov-2017 17:23


Has anyone developed any 4GL to comunicate with voice recognition devices? The device would be a warehouse barcode scanner with voice recognition. Any information on how this environment works would be appreciated.

Thanks, Anthony.

All Replies

Posted by James Palmer on 06-Nov-2017 03:20


We have done so. Unfortunately I'm not too close to the project and it's a project we make quite a lot of money with so I'm not at liberty to divulge details. We've developed an API that runs via AppServer that the handheld devices key into. I do know that the one thing to look out for is the speed of the API. Too slow, and the users get impatient, too quick, and they miss prompts and have to ask the device to repeat. That then slows them down again. Also, make sure the wifi signal is very good in every location they will be using the devices. You don't want to lose connection in the middle of a pick!

Sorry I can't be more help other than to say it's possible to do.

Posted by uktiffy on 06-Nov-2017 03:37

It depends on the capabilities of the machine.  We developed a voice application where the logic was done on the machine based on data that had been pushed from an AppServer.  This meant that a connection wasn't always required by the application so long as it got a connection at some point.  This also allowed the voice machine to go at its own pace so to speak.  However, a good wifi coverage will make matters a lot better should you have to cope with the unexpected.

Sorry, this doesn't answer your question but the development was about 15 years ago and I've moved since then.  But as James says it is very possible!

Posted by marian.edu on 06-Nov-2017 06:34

I doubt there is any standard there so it all depends on what the scanner offers in term of integration/SDK, but the ‘conversation’ is probably started by the scanner and this will lead it till it ends so on 4GL side you’ll probably just have to implement some web services (rest/soap) to respond back.

Who do you expect the two words to communicate, what will be the difference between a barcode scanner and a voice recognition enabled one as far as the 4GL app is concerned? Do you expect those to get involved in some sort of conversation like with one of the assistants out there?

Marian Edu

Acorn IT 
+40 740 036 212

Posted by Scott.Winkeler on 06-Nov-2017 11:28

We use Vocollect headsets in our part inspection tunnel.  The users call out defects on the painted parts in their headset.  The Vocollect system drops a text file to our server and we have a progress server side application that creates database records off of the text files being created.  

We also have a WMS system that our forklift drivers scan containers and racks of parts to lanes in the warehouses but all of that is done through scanners and barcodes.  We haven't had a need to introduce any voice recognition into the WMS applications yet.  

Posted by bbrennan on 07-Nov-2017 07:46

What hardware and middleware are you using.  As suggested, if this is still an open issue, I would recommend you consider Vocollect Voice by Honeywell.  Paired with Honeywell mobile computers, it provides a very performant and rugged solution.
I will also second the idea of having voice quality wireless coverage in very corner of the facility.  Make sure to have a post installation survey completed that details universal coverage.  This is not using your iPhone to walk around and look at bar strength.  Find a vendor equipped with products like Air Magnet or Ekahau that will document the coverages, speeds and noise in the environment.  Depending on where you are located, we might have some contacts for your consideration.
Brian Preece at Ypsilon (http://www.ypsilonsoftware.co.uk) would also be a resource to contact.  He and I have discussed his success in this area a number of times over the years.
Bob Brennan
Integrated Manufacturing Systems, Inc.

This thread is closed