MyVoice is a group of customisable, natural language speech applications that allows users to interact with the MFD in a conversational fashion to achieve tasks. Now, everyone is probably thinking "but speech recognition is unreliable, it will never work" or "I'll feel silly talking to a machine". We are aware of this...

Current attempts to implement Voice for MFD control have centred around accessibility. While MyVoice is obviously a very good accessibility aid, that was not our only design goal. We wanted a Voice solution that anyone would want to use.
We paid great attention to the following:

1. MFDs or any embedded speech solution will not be powerful enough to deliver a natural conversation without making it non-viable on cost. This is why existing solutions offer limited menu choices: "Choose 1 for X, 2 for Y"  and that was not what we wanted. The unique architecture of m3i allows us to control the MFD over the network, so the obvious answer for us was to move the voice processing and the speech application away from the MFD onto a server capable of processing it, which in turn tells the machine what you want and tells you what the machine is doing.
2. People feel awkward talking to machines and in a noisy office this is probably not possible both in terms of background noise and the annoyance to other office users caused by people 'tlaking' to the MFD.
   The simple solution to this was to use a device that people are extremely comfortable speaking to - a phone. Unfortunately this doesn't help with problem 3:
3. To have a conversation with the MFD you have to feel a connection with it (calm down, you don't need to take it to dinner). However natural the voice application may be, you don't want to feel like you are talking to some remote proxy that in turn is relaying your requests. We have tried to make the handset feel like part of the MFD. It's the right colour and looks comfortable (don't worry about the photo it's a mock-up). It has no dial-pad. You don't dial the MFD, you just pick it up and talk. We give the MFD a friendly name, a personality almost. All this is geared around making you comfortable.
4. Voice is not the whole solution. By this we mean that speech, by its nature, is low bandwidth. It's much faster than typing but much slower than visual input. So the answer is to use a combination of User Interfaces. Verbally issue quick instructions like: "Scan to Email, A4, Black & White, Send to Bob Jones". This will be (briefly) verbally confirmed, but then the detail and the progress can also be visually confirmed on the MFD screen or on a web browser if desired.

Key Features

• Use our integrated handset or any industry standard SIP phone/handset. If you already have a VoIP system, you can integrate with that or you can use our handset. You can even use soft SIP phones on mobile phones so you can talk to the MFD over WiFi if you prefer. The options are limitless.
• Voice processing will be available as a hosted service or as a local server option. Your choice will depend upon your environment and requirements.
• Ability to integrate other external information or support functions. A user can optionally report problems, get help or even speak to a real person if required.
• Like all m3i Components MyVoice works collaboratively, so you can just use voice appropriately and not necessarily all the time.
• It's application specific but we have built in the ability for the Speech Application to be able to dynamically alter its verbosity based upon how competent it believes you to be. The first time you use it, you will find it explains more and gives you more time. If you start taking shortcuts, it will learn to keep quiet unless you seem lost. Clever?
• Accessibility: Section 508, Disability Discrimination Act etc.
• Just like the other m3i Components, MyVoice is extensible. When it finds itself in a Copy Accounting environment it will incorporate cost and balance information into the conversation and allow user authentication via voice. It could easily accept balance top-ups from users if required.