Since each person 's voice is different, the software of the SpeakUp Click module will let me tailor the vocabulary [any language], and increase the chance of word recognition by having at least two variations of each command word in the data base. Debugging of the setup is easier since I can 'evesdrop' on the data the unit sends to the wheelchair controller microprocessor.

A long time ago, I developed a philosophy that pre-dated Hackaday, that I described as "why re-invent the wheel, when you can steal a hubcap". So rather than attempt to develop a voice recognition module's hardware and software, I felt it was faster and [for time reasons] cheaper to use an off the shelf module, and take advantage of it's hardware and support software.