Information Technology (IT) is a core business capability for companies. Intel IT delivers the critical information
solutions needed to keep Intel running and growing. Over the last few years, Intel has focused on IT innovation, with
an exploratory research direction called the Proactive Enterprise. Our aim is to shift current enterprise computing
architectures and usage models from a mostly reactive and human-centered state toward one that anticipates our needs
and is proactive, integrated, and human-supervised. Our four-layer framework (shown below) represents emerging research
in enterprise infrastructure, information management, business processes and user application systems.
Collaborative Computing
Our research goals involve improving global team productivity to support decision making and reduce rework and time-to-market.
In 2004, we provided strategic direction for digitally enhancing the workforce environment, and worked with Intel's Digital Office program on the Collaboration Concept Vision.
In 2005, we plan to establish a Collaboratory, a physical lab environment, and a virtual ecosystem to attract and catalyze research on next generation large-scale collaboration.
Smart Diagnostics
To control maintenance costs and minimize unexpected downtime, we are developing models to streamline diagnosis and repair of manufacturing tools using Bayes networks.
Although our focus is on factory equipment troubleshooting, we expect our research to benefit any complex diagnostic domain.
PlanetLab for the Enterprise
PlanetLab is a testbed for developing and deploying planetary-scale services and an improved Internet architecture.
We are investigating how PlanetLab can be configured for secure use by enterprises and how it can address total-cost-of-ownership challenges for planetary-scale applications.
In 2004, we published a vision for applying PlanetLab and its core technology to manage enterprise systems.
Data Mining in Semiconductor Manufacturing
We developed a proof-of-concept data mining technique for improving factory planning and demand management.
We improved our prediction accuracy over traditional methods and began piloting the data mining technique in 2004.
Web Trust
We explored ways to make the Web a more "trusted" and preferred medium of choice.
In 2004, we added over 30 new intranet assistance wizards, reducing trouble calls significantly, increasing customer satisfaction, and realizing substantial cost savings.
High-volume Manufacturing with System for Human Activity Recognition
We are optimizing factory equipment maintenance processes with Human Activity Recognition, by training computers to anticipate our needs and produce responses without intervention.
We expect our research to influence Best Known Method (BKM) identification, training and computer assistance.
Demand Generation
We've developed structured assessments and models that increase Intel major product transition success by integrating organizational domain knowledge.
We piloted these methods in 2004 and will continue working with our partner groups to integrate them into standard processes in 2005.
Policy-Based Security Management Architecture
We created an automated and adaptive management architecture that coordinates network security solutions and enables management of common security policies across a heterogeneous enterprise network.
In 2004, we started a proof-of-concept project to use this architecture to improve efficiency for security policy management and reduce network response time to ever-changing attacks.
Collaborative Decision Making in a Federated Enterprise
Supporting collaboration throughout the value chain, we are exploring an integrated process and technology framework as a two tier strategy.
We are applying the Value Chain Operations Reference model and Federated Enterprise Reference Architecture to enable both tiers.
