Ubiquitous computing (ubicomp) is a vision of hardware/software systems that pervade our environment, doing what we want without our continual direction. For example, consider a motorway for driverless vehicles. One of the UK Grand Challenges for Computing Research addresses ubicomp not only from the visionary angle, but also from the angle of design principles and theories that will support it.
Ubicomp systems will exceed those that we know by orders of magnitude. There seems little chance of extrapolating existing methods of software production to cope with them. Ubicomp offers a challenge and an opportunity to develop a new computer science that interweaves three ingredients -- vision, design and theory -- much more intimately than ever before. We need this, to be confident in unleashing ubicomp systems.
That's the Grand Challenge; the lecture will explore how to address it. One idea is via foothill projects, which address modest aims but try to combine segments of vision, design and theory. I shall illustrate this with a mathematical model of mobility that links the physical space of sensors and buildings with the virtual space of software processes. If nothing else, this exercise shows that we have a long way to go if we want to understand mobile ubiquitous systems.