Roboticist Robert Riener at the Swiss Federal Institute of Technology (ETH) in Zurich has developed a new robotic bed for sleep research and therapy. The Somnomat uses a system of cables attached to the posts of a suspended bed to move the bed in whatever sleep-inducing pattern the researcher/insomniac wishes to test/try to fall asleep with. He reported it during a recent conference in Lausanne on "Engineering Life" that I also gave a talk at. A short video of it is here. There has been a something of a renaissance of interest in sleep. It is, after all, something we engage in daily for about a third of our lives without tiring of the activity. There are a lot of consequences of not doing it well. For one thing, we can become an emotional wreck when we don't get enough sleep. In healthy people that are sleep deprived, the part of the brain that regulates emotion becomes hyperactive, in a fashion similar to what is seen in depressed individuals. For me it's one of the most reliable side effects of not getting enough -- suddenly events or things people say that would otherwise be neutral have a much higher chance of affecting me emotionally. Sleep appears to be essential for remembering things we learn over the long haul. Most recently, there's evidence that lack of proper sleep before the age of five can significantly increase the chances of being obese later in life.
For all its importance to our well-being, you might think we would have a handle on everyday observations of what makes people more likely to fall asleep. For example, rocking a baby has been known to help put babies asleep probably even before we had the language to express this. There is a similar effect of rhythmic movement on adults. Yet, why it is we find rhythmic movement soporific is not currently known. Surely knowing more would be very helpful: how much movement is best for falling asleep? Should it just be for falling asleep, or continue after that? What is the best pattern of movement? How does this mechanical approach compare to other approaches, such as drugs? These and other questions can be addressed with the somnomat. If it turns out to be as beneficial as generations of experience of rocking our children asleep would suggest, then getting these details right could be immensely useful for designing a new kind of automatically moving bed that helps people fall asleep. Some updates to our favorite lullabies may be needed... Rock-a-bye baby, on the robotic treetop, When the servomotors turn, the cradle will rock, When the robot breaks, the cradle will fall, And down will come baby, cradle and all.