Designing Future BCIs: Beyond the Bit Rate

The scope of this chapter is limited to applications where a Brain–Computer Interface (BCI) is used as an explicit interaction technique. In other words, we refer here to BCI as input which is voluntarily controlled by the user, rather than as an implicit interaction as in for mental or cognitive state monitoring. Designing applications using BCI as an explicit input technique for users with severe disability depends on understanding the control signals and how users can interact with systems using these controls. Although designing for able-bodied users has a different set of challenges, the BCI has to “add value” in both cases. Over the past 20 years of BCI research and design, the basic control functions have been realized by the collaboration of engineers, psychologists, machine learners and end users. These basic functions provide us with the freedom to design future BCI applications which are reliable in long-term use, easy to learn and set up, aesthetically pleasing, and have the potential to improve the lives of their users. BCI can be thought of as an input technology which takes properties of other emerging input technologies to the extreme. The term “extreme” is used because the BCI interaction is much slower, noisier and more error-prone compared to other input devices, and lacks proprioceptive feedback. Because of these unusual characteristics, a theoretical framework which successfully analyses current BCI systems provides a springboard for developing and refining theories and practices within Human Computer Interaction (HCI). Although still important, research in