In the past few years there have been great improvements in voice-to-text technology. Google supports it by default on their Android handsets, plus other companies like Vlingo have similar offerings. The accuracy and reliability is constantly improving.
What I want to know is if a technology like this could be useful to end-users. We currently have dedicated speech terminals which recognize a small set of commands. This wouldn't be a replacement of that (yet), but would allow a user to perform general text entry by speaking at the terminal.
My first thought is when the terminals are docked in a vehicle, if you need to send an email, initiate a phone call,... you shouldn't have to take your eyes and hands away from the task of driving. On the flip side, what about a text-to-speech functionality? So if you did receive a message, the terminal could read it out to you.
Thoughts?
Voice to Text might be useful for reporting tasks done by mobile workers. We've heard of some reports being started using drop down fields to select what will be in the report outline, and then polishing up the report later on away from the jobsite. Being able to dictate your reports could streamline this process, if it's reasonably accurate.
I think that would be very usefulf if we could get voice commands to trigger events on a terminal. There are only so many keys that can be re-mapped on a handheld. Could get a voice command to trigger a scanner, or to take a picture for instance?
Its an interesting area - could you overlay the voice to text engine so that it can input into any application that is being run on the screen - effectively whereever the cursor lies?
In warehouse, retail applications (indoor basically) where voice directed applications sit the benefits of this is that you are predominantly not needing to use a terminals other functions very often - only by exception or to add speed in certain areas. So not sure if there is a huge benefit here.
Perhaps an application area for this would be more in mobile devices where remote engineers are needing to make notes on a customer/job record of the work they are doing. They could just dictate the information and it gets converted to text in an input field.
I know they do this in hospitals at the minute but they have to sit at a PC and dictate the notes back - its not done at the point of note capture. I believe they use Dragon