1. Gaetano Borriello Department of CS&E University of Washington The Portolano Expedition in Invisible Computing portolano.cs.washington.edu

2. 14 September 2000Seattle SAGE Group2 Context  Today, only 2% of microprocessors are used in what we think of as “computers”  This percentage will continue to decline  Five major DARPA sites focused on systems involving the other 98% -- on “invisible computing”:  UW  Berkeley  MIT  Carnegie Mellon  Georgia Tech  UW and Berkeley are closely collaborative, closely tied to Intel

3. 14 September 2000Seattle SAGE Group3 Principal Themes  Invisibility  not enough to be mobile, pervasive, ubiquitous, etc.  user’s attention is the valuable resource  minimize user configuration/maintenance/interaction  robust, reliable, safe, and trustworthy  devices, middle-ware, and “applications”  services  Active fabric  plug-and-play, discovery, composability  data-centric, heterogeneous, active networking  data and code mobility  self-organizing, self-updating, self-monitoring systems  active databases and information management  External user community

4. 14 September 2000Seattle SAGE Group4 World View  What will be:  lots of task-specific devices  rich connections to physical world  world-wide information/computation utilities  ever-increasing computation, storage, and bandwidth  computing/communication capabilities in everything  What won’t be:  unlimited power sources  homogeneous connectivity  continuous connectivity  increased user mind-share devoted to computing concerns

5. 14 September 2000Seattle SAGE Group5 Expedition Goals  Connecting the physical world to the world-wide information fabric  instrument the environment: sensors, locators, actuators  universal plug-and-play at all levels: devices to services  optimize for power: computation partitioning, comm. opt.  intermittent communication: new networking strategies  Get computers out of the way  don’t interfere with user’s tasks  diverse task-specific devices with optimized form-factors  wide range of input/output modalities  Robust, trustworthy services  high-productivity software development  self-organizing, active middleware, maintenance, monitoring  higher-level, meaningful services

6. 14 September 2000Seattle SAGE Group6 Application domains (and collaborators)  Labscape (UW Cell Systems Initiative)  instrumentation of the workplace  capture experiments as they happen (build the database by doing the work)  data mining to support investigations/recollections  Personal devices/networks (Microsoft Research)  body-area networking (RF and skin)  borrowable, scrap devices  Here-to-there (Intel, Ford, Xerox)  in-building/in-room location tracking  continuous access to data and multi-media streams  management of personal/group devices and services

7. 14 September 2000Seattle SAGE Group7 Research Themes  Low-power intermittently connected devices  Intentional user interfaces  Data-centric networking  Self-organizing information systems  Seamless data/information management services active fabric devices

8. 14 September 2000Seattle SAGE Group8 publication LabScape - one of our driver applications  Biology is a hard science with a soft infrastructure  capture and use of knowledge is key  from loosely connected to highly integrated collaboration  invisible infrastructure for building knowledge base Interpret ExperimentHypothesize Interpret ExperimentHypothesize Descriptive Model Experiment Manager knowledge base

9. 14 September 2000Seattle SAGE Group9 Event Capture in Labscape