Auditory feedback devices are a mixed bag. There's the ever-popular speech-easy device. Works for some people, doesn't for others, and often, even when it does work, the user gets used to it eventually and it stops working. $5k down the drain. Robots to make phone calls for us won't help. That just enables us in avoiding situations. I'm not sure how a robot teaching people how to enunciate wounds would help, we can already make the sounds we need to make, we just get stuck on them. However, a lot of the time, especially on consonant sounds, people who stutter put way too much pressure and tension on the muscles needed to make those sounds, really emphasize them. Perhaps a visual representation of light touches could help.