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.
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.
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!
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.