Upload
timothy-davidson
View
212
Download
0
Embed Size (px)
Citation preview
Computing and Communications Technologies:
Quo Vadis ?
Athens, 20 May 2001
Horst ForsterEuropean Commission
ATH0
DistributedSystems
Control SystemsCognitive Vision
EmbeddedSystems
Optical NetworkingNetwork Management
Distributed ApplicationsService Provision
PervasiveComputing /Networking
Computing and Networking
IBM0
Where do we go from now ?
Computing and embedded software are ubiquitous
Networks and bandwidth are proliferating. . . and growth continues
Cheap miniaturised sensors emerge
IBM1
Implications
Infrastructure supports more functionality
Applications become distributed
Applications are tethered to the physical world
Systems are increasing in scale and complexity
ATH2
Implications:Infrastructure Evolution
1996-2000 Static optical networkingData prevails over voice Cellular telephony on the rise
2001-2005 Dynamic optical networkingData-centric networksCellular telephony meets Internet
2006-2010 All-optical networking ?and beyond Towards “4th G” ?
ATH3
Challenges: Infrastructure
Transformation
. . . from connected networks and computers
. . . to service access and delivery environment• adapt automatically and dynamically• distributed (and possibly mobile) demands• satisfy all aspects of the user’s computing,
communication, storage and service functionality
CONTROL
ATH5
Challenges: Applications
Transformation
. . . from processing and delivery of data
. . . to interaction with many data sources, dynamically changing data-sets and tasks
• augment human capabilities in analysis, interpretation and decision-making
• computation cannot always prevail• master scale and complexity in an intelligent
way
COGNITIONATH6
In short
As computing and communications
become ubiquitous, control and
cognition will rank as key enabling
technologies
ATH7
Cognitive vision systemsExamples of projects
COGVIS
-provides methods to enable task-oriented categorisation and recognition of objects and events
-system:‘understands’ - recognition/categorisation through association‘knows’ - memory as a basis for information representation/maintenance‘learns’ - allows operation beyond initial specifications
-a prototype integrating methods will interpret user actions and perform tasks,e.g. fetch and deliver objects in realistic settings
COGVISYS
-builds a vision system that can be used in a wider variety of vields
-system: self-adapts at level of perceptionuses explicit knowledge base at level of reasoning
-demonstrators will translate visual information into textual descriptione.g. for traffic surveillance, sign language interpretation, video annotation
IBM2
Pervasive ComputingExample
O Z O N E
nomadic use of technology
natural interfaces
pervasive connectivity; location & situation awareness
dynamic configuration of computing elements and services
high performance (low power) hw/sw platforms
low cost
IBM3