Input Research at Microsoft: Sensors, Mice, and Keyboards, Oh My!

Ken Hinckley, Microsoft Research
   kenh@microsoft.com

Seminar on People, Computers, and Design
Stanford University March 1, 2002

It has become almost fashionable for HCI and technology prognosticators to predict the rise of sensor-rich, smart environments and devices accompanied by the imminent demise of the mouse and keyboard. But then why is it that, some 35 years after the invention of the mouse, it is now used more widely than ever? Why are keyboards still being used well over 100 years after the introduction of their predecessor, the typewriter? What is the future of our trusty friends, the mouse and keyboard? What are some of the all-time dumbest strategies that have been tried to improve input devices? :-)

At Microsoft Research, we believe that sensors and smart devices will co-exist with traditional mice and keyboards. We are excited by the possibility of making mobile devices just a little bit smarter by adding inexpensive, low-power sensors for contextual awareness, allowing a device that is smart enough to turn itself on if you pick it up, or switch to landscape display mode if you hold the screen sideways, for example. These type of sensors are a reality now and just waiting for innovative user interface applications. But the same general strategy can be applied to traditional devices such as mice by making them touch-sensitive, for example, so that wireless mice can power themselves off when not in use, as done by the Microsoft Wireless IntelliMouse Explorer.

We also believe it is important to research problems such as document navigation (scrolling), and improve that experience for millions of people using mice and keyboards. For example, I will argue that all studies of scrolling in the HCI literature suffer a common fundamental flaw, which can be addressed by applying Fitts' Law to the problem. This insight has led us to more clever real-time algorithms for handling the IntelliMouse wheel sensor's data, resulting in significantly better scrolling performance. I will also discuss the design of the Microsoft Office Keyboard, an innovative new commercial keyboard designed to encourage bimanual division of labor between the left and right hands.

Ken Hinckley is a Research Scientist at Microsoft Research. Ken's research interests include human-computer interaction, input devices & interaction techniques, mobile devices, and experimental studies of human abilities. Ken received the UIST 2000 best paper award for his work on Sensing Techniques for Mobile Devices, along with co-authors Jeff Pierce, Mike Sinclair, and Eric Horvitz. Several technologies that Ken has worked on are now commercially available in the Microsoft Wireless Explorer mouse, Office Keyboard, and associated software. In his spare time Ken likes to go adventuring in the Cascade Mountains and is an avid basketball player and watercolor painter.

View this talk on line at CS547 on Stanford OnLine

Titles and abstracts for all years are available by year and by speaker.

For more information about HCI at Stanford see