This area includes several methodologies underlying the area of Information Processing and Networking, namely, information networking, language technology, software engineering, and information infrastructures security and dependability. This area involves partnerships of several CMU Departments and Research Centers and several Portuguese Universities and Research Institutions.

Today’s world of rapidly changing software technology underscores the need for software engineering research and education to deal with new methods, tools, platforms, user expectations, and software markets that address the large scale problems that dominate current relevant applications. The programs envisaged in this CMU-Portugal cooperation target the further development of the educational and research programs at Portuguese Universities in IT and software engineering, namely, with Universidade de Coimbra, through its Faculdade de Ciências e Tecnologia, among other Portuguese schools. These programs hold the potential for a significant impact on traditional large companies consumers and producers of software (e.g., Portugal Telecom) and its associates but also on the nascent Portuguese software industry. In the last decade, several software related companies (e.g., Critical Software SA, Wit-Software, Florde- Utopia, Insert-Solutions, Netvita, Bookmark, to name a few) have initiated their activity around the University of Coimbra, their Centro de Informática e Sistemas da Universidade de Coimbra (CISUC), and Instituto Pedro Nunes.

The CMU-Portugal Program will include an advanced educational program, the Professional Master of Software Engineering offered in partnership by the University of Coimbra and ISRI at CMU, as well as research projects in software engineering. Other potential partners in Portugal, namely for research activities, will be defined during the initial phase of the Program.

Besides the involvement with private corporations, the CMU-Portugal Program and ICTI will establish partnerships with other Institutions and Agencies. In particular, two governmental agencies have been recognized as main partners of ICTI, namely: “UMIC - The Agency for the Knowledge Society”; and “FCCN – Fundação para a Computação Científica Nacional”, the national foundation for scientific computation.

Other national agencies and governmental departments in Portugal are expected to be involved in the program.

Dependability and security are very important fields in information technology. Dependability attempts to keep systems working correctly despite the occurrence of accidental faults or defects, while security addresses problems arising from malicious hazards, attacks, or intrusions. Because systems are increasingly so complex, dynamic, and interdependent, it is no longer possible to consider these two problems, security and dependability, independently. In a modern perspective, security and dependability are concerned with both information and infrastructure. Of particular interest, are critical information infrastructures, their pervasive interconnection, and the progressive intertwining of “normal” and embedded systems.

The Internet has been transformed by the proliferation of embedded, inconspicuous, and often mobile devices that cluster and un-cluster in ad-hoc fashion. Many are connected to physical artifacts, like in smart homes or ambient intelligence. This reality includes wifi-enabled small computers, wireless sensor and actuator devices, network-enabled embedded gadgets that are present in every day life, but also large scale systems like the telecom network, or the power grid that are permeated with computers that introduce digital control. This new reality poses major challenges from a security, dependability, and resiliency points of view against threats that are not known à priori. These threats include accidental hazards or faults and malicious attacks or intrusions. Given the complexity of these modern large scale infrastructures, the educational and research program on security and dependability will pursue the development of rigorous design methodologies for fault and intrusion prevention, tolerance, and detection.

The proposed educational and research program will address both security and dependability on an equal footing. The collaboration between CMU and Portuguese Universities, in particular, as envisaged with Faculdade de Ciências da Universidade de Lisboa, FCUL, involves coordinated education and research components to address dependability and security in these critical information infrastructures. Of significance is an advanced degree, a Professional Master of Science in Information Technology: Information Security, to be offered in partnership by FCUL and FCUC, among other potential partners in Portugal, and INI at CMU. This program is complemented by a PhD program offered in partnership between FCUL and FCUC, among others, and the Departments of ECE and CS at CMU, and research projects in the areas of security and dependability. Other potential partners in Portugal, namely for research activities, will be defined during the initial phase of the Program. In particular “FCCN – Fundação para a Computação Científica Nacional,” the national foundation for scientific computation, will actively participate in this area, working in close collaboration with university groups.

Information networking is an area of great impact. The CMUPortugal Program will partner Carnegie Mellon faculty drawn from CyLab, CenSCIR, ECE, and CS and several Portuguese research groups, including those at Universidade de Aveiro, INESCID, IT, and Universidade de Coimbra. There will also be a strong involvement with companies in the Portugal Telecom group. In a relatively short time, the Internet has evolved from a small research network used by researchers into a critical infrastructure that delivers a wide variety of services to hundreds of millions of users. Looking forward, we see a number of trends that are likely to cause a similarly dramatic transformation in the next ten to fifteen years. First, while the Internet initially connected fixed, wired, computers, current trends suggest that in the near future, the vast majority of users will use wireless, mobile devices to access Internet services. These personal devices will be complemented by large numbers of non-computational devices, including sensors, actuators, and I/O devices, most of which will also be wireless. This means that wireless will be pervasive as an access network technology. At the same time, the service infrastructure is evolving from simple client-server applications into a sophisticated, highly distributed, highly resilient software platform that delivers personalized services to users. These trends suggest a number of important research areas in information networking. For brevity, we illustrate by focusing on research in wireless networking and the delivery of electronic and pervasive computing services.

The pervasiveness of wireless networks creates a number of challenges. A first challenge is at the network architecture level, where mobility, device heterogeneity, and variable network properties will require new architectural paradigms for access networks that can maintain high service quality for mobile users. These will include new proxy architectures, security and authentication mechanisms, naming and addressing mechanisms, and bandwidth provisioning techniques. At a lower level, the high demand for wireless bandwidth will put pressure on a scarce resource, namely the spectrum. This will require improved efficiency at all layers of the wireless protocol stack. This can be achieved by “autonomic” networking techniques that automatically optimize the network based on wireless channel conditions, traffic load, and node properties – manual network configuration and control will be impractical because of network complexity and dynamics. At the same time, architectural solutions such as spectrum-agile networking, in which nodes opportunistically and dynamically use available spectrum, can be used to increase network capacity.

In order to have impact, this wireless networking research agenda will have to be executed in a realistic, forward-looking broader systems context. The research will have to be driven by aggressive applications, such as high quality video stream and interactive games, and highly mobile users, including car-based wireless networks. Results will have to be evaluated using both large testbeds that combine a variety of devices and applications in a realistic way, and emulation techniques that allow a more controlled quantitative evaluation of individual techniques.

Another key challenge in information networks is the development of a scalable infrastructure that can deliver personalized electronic services in a resilient and secure fashion. Examples of such services include information retrieval, e-commerce, and video conferencing, but also more aggressive services such as virtual reality, remote medicine, and interactive games. Three factors combine to make the delivery of such services a very challenging problem: the complexity of services themselves, the dynamics and unreliability of the underlying hardware infrastructure (network, servers, storage), and the variable user demand.

This infrastructure will require a highly modular approach in which services are constructed from basic building blocks. Not only does this maximize the potential for software reuse, but it also opens the door for the automated configuration and optimization of network services using closed-loop control systems. The service infrastructure will include a set of monitoring probes that can quickly detect failures, changes in load, performance problems, and intrusions. The output of these probes feeds into a diagnostic module that can identify the cause of the problem and formulate an automated response. This will require research of both an experimental and formal nature in a variety of areas, including distributed systems, software engineering, formal methods, and security.

Besides the delivery of the above-mentioned “electronic” services, we also envision a dramatic growth in the delivery of “physical” services that leverage the rapid deployment of sensors, actuators, and I/O devices. Examples of such “pervasive computing” services include location- and context-aware services, automated control of the user’s physical environment (heat, light...), and personalized entertainment. The requirements in this area are similar to those mentioned above for the electronic services infrastructure: how can we deliver personalized services in a resilient and secure fashion. However, because of the localized nature of the services and the tight coupling to the physical world, the challenges are different. Questions include: how to manage the large volume of data, how to translate noisy, dynamic sensor data into useful information, how to engineer applications that continuously adapt to the user’s context, and how to maintain privacy and security in this device rich environment.

Addressing these challenges requires an architecture that separates the responsibility of the different players using well-defined interfaces. At the lowest layer, we need techniques for the configuration and management of the sensor networks so they can deliver the sensor data to a set of information services in a timely fashion. The sensor data will be filtered and combined with data from other sources (e.g. calendar and other databases) to create context information that will be tagged with meta-data, e.g. to time stamps, precision, etc. This will effectively create a distributed database of context information that can be used by a variety of applications. If designed right, this architecture and its component can be reused in many environments, e.g. office, home, etc. This pervasive computing infrastructure can be built, deployed, and maintained by commercial service providers, similar to the electronic services infrastructure.

The CMU-Portugal Program includes an advanced degree program, a Professional Master of Science in Information Networking, to be offered in partnership by the University of Aveiro and the INI at CMU. Other potential partners, namely for research activities, should be defined during the initial phase of the Program. A PhD and a research component to be pursued in partnership with the ECE Department at CMU complement this program.

The CMU-Portugal Program on language technology involves a consortium of Portuguese Research Centers and Universities and the Language Technology Institute (LTI) at CMU. The LTI formed about 20 years ago, first as a research center, and then as an academic department in the School of Computer Science, is the leading center in language technologies in the world. The consortium in Portugal, which we will refer to as the L2F consortium, includes the Laboratório de Sistemas de Lingua Falada (L2F) (language to speech) at INESC-ID, IST, the Center of Linguistics of Universidade de Lisboa (CLUL), the group of linguistics of Universidade do Algarve (UALG), and the Center for Human Language Technology and Bioinformatics (HULTIG) of the Universidade of Beira Interior (UBI). In addition, a close cooperation is expected with the LINGUATEC network established in Portugal through FCCN (i.e., the national foundation for scientific computation, through the "Centro de Recursos Distribuído para o Processamento Computacional da Língua Portuguesa"). The collaboration between L2F and CLUL dates back to the early nineties, forming the basis for a truly interdisciplinary cooperation (engineering/linguistics). The cooperation with UALG and HULTIG is much more recent and, despite their much smaller size in terms of human language technologies, is also very active.

There are a number of areas of strong common interest that will be pursued: computer aided language learning (CALL), speech-to-speech machine translation (S2SMT), speech recognition, speech synthesis, dialogue systems, summarization, and topic detection and tracking. In particular, the CMU-Portugal Program will pursue two very important multilingual research projects: in computer aided language learning (CALL); and speech-to-speech machine translation (S2SMT). These projects will involve at least two languages, one of them being Portuguese, the target language for the CALL system to be developed and either the source or target language (or both) for the MR system. The other language is either English or Chinese (Mandarin) or both. Chinese is of particular interest to both parts, because of the existing expertise at LTI with language technologies for Chinese and the great demand from China for products involving Portuguese.

The CALL research project will involve the development of a Portuguese version of the REAP project (Reader-Specific Lexical Practice for Improved Reading Comprehension), currently in progress at LTI at CMU, and research in the associated topics. This research involves a variety of language technologies to help native or non-native students learning to read. Examples include searching for appropriate authentic documents for students to read according to reading level, topic, vocabulary list, and other teacher-specific criteria, with a search engine that finds text passages satisfying very specific lexical constraints, selects materials from an open-corpus, thus satisfying a wide range of student interests and classroom needs, and models an individual's degree of acquisition and fluency for each word in a constantly-expanding lexicon so as to provide student-specific practice and remediation. The challenges posed by such a system enable research on a wide range of very difficult reading comprehension topics.

The speech-to-speech machine translation (S2SMT) research area, one of the most strategically relevant areas for the Portuguese consortium, will investigate a Portuguese-to- English/Chinese translation system (or vice-versa) that deals with one of the two challenges of current S2SMT systems: the need for disfluency removal on the speech input side, or the inadequacy of current translation systems to produce text that is fit for a synthesizer, i.e., that can be read in a naturally sounding way. We will investigate the use of statistical based machine translation approaches, but also a hybrid approach, developed by LTI, that starts from a relatively small parallel elicitation corpus and uses rule induction. Research in S2SMT is crucially dependent on several core technologies from speech recognition to machine translation, to text-to-speech synthesis, including voice morphing.

These two projects provide a focus for the proposed research; through them the collaboration will explore the main core areas in language technology.

The CMU-Portugal Program includes a PhD degree component offered in partnership between IST and LTI at CMU, and research projects in language technologies between the L2F consortium and LTI. Other potential partners in Portugal will be defined during the initial phase of the Program.

The main goal of the partnership is to contribute to promoting the development of the Portuguese scientific and technological capabilities necessary for both the operation and performance assessment of public risk minimization systems. This will rely on an integrated research and educational (PhD) program in the area of Critical Infrastructures and Risk Assessment, specially focused on Networked Sensor, Communication, and Decision Systems, which are important and very actual topics in Information and Communication Technologies (ICT). Those topics include the sub-areas of wireless sensor networks, HW platforms and sensing technologies, software and middleware, decision making, actuation, security, privacy, applications, and risk assessment.

The partnership has been launched in Portugal with ISR-Lisbon, INESC-ID, CISTER/ISEP, and ISQ, but it is expected to bring together other leading Portuguese institutions. The following CMU Units are involved: Center for Sensed Critical Infrastructure Research (CenSCIR), Electrical and Computer Engineering (ECE) Department, Computer Science (CS) Department, Engineering and Public Policy (EPP) Department, and Tepper, the CMU Business School.

The participant researchers, from the Portuguese partners and from CMU, form a multidisciplinary team that cover all the sub-areas identified in the CIRA area.

This is a collaborative scientific program in the general area of CIRA, promoting joint research initiatives, integrating the capabilities of the Portuguese research units and of CMU, and including a dual doctoral program in ECE to be granted by IST – UTL and CMU, the main focus of which being on Networked Sensor, Communication, and Decision Systems. Other potential partner universities in Portugal will be defined during the initial phase of the Program. The research objectives in this collaboration cover the sub-areas of wireless sensor networks, HW platforms and sensing technologies, software and middleware, decision making, actuation, security, privacy, applications, and risk assessment, and other technological/ scientific related topics such as: static and mobile sensors; data, image, and video processing; distributed information fusion and perception; computer vision; optimization; cooperative planning and decision making. This research activity will be strongly connected to the doctoral program in ECE, although in some specific cases it may support students involved in other doctoral studies directly related with the objectives of this partnership.

The doctoral program is designed to take up to five years to complete. Students are required to fulfill course requirements covering fundamentals and applied topics in ECE and in other relevant areas, such as risk analysis and decision support methods. Typically, each student will spend the first year in Portugal taking the courses on fundamentals. The next 1½ /2 years are spent at CMU to complete the course requirements and the qualifying examination. The final 1½ /2 years are spent in Portugal. Since the beginning, the student will be involved in research activities related with the theme of his/her thesis.

Technological change and innovation are at the heart of regional and national economic growth and firm performance. Thus, while Portugal needs to develop new knowledge and technologies, it is critical for the nation to be able to apply them in the market for the successful creation of new products, processes, and services. The development of strategies and policies to guide such innovative activity at the national, regional and organizational levels is thus an imperative for productivity and employment growth. This is particularly relevant in the areas of information and communication technologies. ICTs are at the core of the innovation process, not only in industries such as computers or software, but across all sectors, as organizations in every area of activity adopt ever more sophisticated ICT systems.

This innovation process is characterized by complex phenomena that have science or technology at their core, but where social, economic, and organizational forces also play critical roles. Thus, the collaboration program integrates a set of research and education efforts that aim to complement and extend the technical effort, focusing on the development of the analytical lenses and tools needed to study and understand the critical social and economic dimensions of innovation phenomena, especially those associated with ICTs. This collaboration between Carnegie Mellon and research institutions in Portugal, including Universidade Técnica de Lisboa (UTL), Universidade Católica Portuguesa (UCP) and University of Minho (UM), will pursue research and education of highly qualified people, capable of leading research, education and policy with this integrated perspective. Application to improve understanding of public actions and to help designing public policies for the information society will be established through UMIC.

The main subareas encompassed in the program are: a dual Ph.D. program that aims at providing students with substantive skills for academic research and for high level business/corporate strategy and public policy formulation in technological change and innovation dynamics; and a research program in engineering and public policy. The latter subarea has three main focuses, namely telecommunications management and policy, public policy for power grids, and engineering and policy applied to the software industry.

Technological change and innovation are at the heart of regional and national economic growth and firm performance. Thus, while Portugal needs to develop new knowledge and technologies, it is critical for the nation to be able to apply them in the market for the successful creation of new products, processes, and services. The development of strategies and policies to guide such innovative activity at the national, regional and organizational levels is thus an imperative for productivity and employment growth. This is particularly relevant in the areas of information and communication technologies. ICTs are at the core of the innovation process, not only in industries such as computers or software, but across all sectors, as organizations in every area of activity adopt ever more sophisticated ICT systems.

This innovation process is characterized by complex phenomena that have science or technology at their core, but where social, economic, and organizational forces also play critical roles. Thus, the collaboration program integrates a set of research and education efforts that aim to complement and extend the technical effort, focusing on the development of the analytical lenses and tools needed to study and understand the critical social and economic dimensions of innovation phenomena, especially those associated with ICTs. This collaboration between Carnegie Mellon and research institutions in Portugal, including Universidade Técnica de Lisboa (UTL), Universidade Católica Portuguesa (UCP) and University of Minho (UM), will pursue research and education of highly qualified people, capable of leading research, education and policy with this integrated perspective. Application to improve understanding of public actions and to help designing public policies for the information society will be established through UMIC.

Innovation and technical change are nowadays the main engines of economic and social development. Thus, the development of strategies and policies to guide innovative activity in countries, regions and organizations is essential for the re-structuring and renewal of market economies, an imperative challenge for productivity and employment to grow. As the process of change accelerates, entrepreneurship is also becoming increasingly prominent. As a result, achieving excellence in education and research in the areas of technological innovation and policy has become essential for the competitiveness of firms, regions and countries.

The foundation to this collaboration is a dual Ph.D. program proposed by CMU, UCP and UTL. The program aims at providing students with substantive skills for academic research and for high level business/corporate strategy and public policy formulation. The target audiences for this program are post-graduate students from engineering, technology, economics, business, and social sciences willing to pursue a career in academia and/or strategic leadership and policy-making in industry or in government. The program should be aiming to attract students from all over the world, with particular emphasis on Europe and developing countries.

The education of highly qualified people in the aforementioned areas should be a priority for Portugal. The Portuguese model of economic development is at a crossroads. The productivity gains arising from public investment in physical infrastructure and development of financial services and real estate markets seem to be tailing off. The transition towards a knowledge-based, entrepreneurial model of development requires the education of high quality human resources that can teach, research and work at high level private sector and public administration positions in technology commercialization and the strategic management of technology. Through their future role as educators and leading experts in these areas, these human resources could significantly influence the development of Portugal. Moreover, these topics have gained increasing importance in Europe and developing countries during the last decade. Thus, the establishment of a strong collaboration with Carnegie Mellon University in these areas has the potential to give Portugal a specialized academic structure that would figure at the top of European higher education institutions, attracting high quality human resources from a variety of geographic origins.

This collaboration is also to be considered as an integral part of the overall association platform between Carnegie Mellon University and Portugal centered in the areas of information and communication technologies (ICTs). Portugal needs to go beyond the development of new knowledge and technologies in these areas, and also lean to effectively apply them in the market, either through established firms or start-ups. The proposed collaboration in Technology, Innovation, and Policy will help promote strategic visions for public and private ventures and develop the necessary skills to help developing partnerships for innovation at an international level in these areas.

This subarea considers three main themes for research. First, it considers telecommunications management and policy. This is because digital convergence is dramatically changing the technology, services and competitive environment faced by telecommunications carriers around the world. Portugal is no exception. Separate infrastructures tightly tied to specific services, such as telephone and cable, are giving way to generic IP based infrastructures providing a wide gamut of services, including voice, data and video. This transformation impacts industry structure, competition and regulatory policy. Critical questions range from “Is the diffusion of broadband best realized by a monopoly provider of infrastructure, or should telephone, cable and wireless based infrastructures be provided by independent entities, each competing for the customer?” to “Should municipalities take responsibility for building new wireless or fiber-based infrastructures in the same way that they build roads or sewer systems, leaving private sector companies to compete only in the provision of services?”

This research, proposed by EPP at CMU in collaboration with IST/UTL and UMIC will focus on two main issues. The first concerns questions of industry structure and regulation. Portugal, as elsewhere in the EU, must decide whether and how traditional open access policies should be carried forward onto new Fiber To The Home (FTTH) networks. The decision rests on choosing the appropriate balance between incentives for investment in infrastructure and facilitating innovative entry and competition in the delivery of services. The second issue concerns the role of individual communities in promoting broadband diffusion and adoption. Cities and towns can take a range of actions varying from direct investment in infrastructure, such as conduits or fiber, to demand stimulation through education and outreach programs. At the same time the national government is wrestling with how to support and orient municipal efforts. Comparing efforts currently underway or proposed in Portugal and in the rest of Europe with U.S. experience in comparable small towns and rural communities can provide valuable insight.

Second, this subarea will also consider “Public Policy for Power Grids”. The development and management of electric power grids and advanced low-carbon uses of coal, other fossil fuels, and the integration of intermittent renewables are key problems that will shape the future development of the electricity industry. This research collaboration between Carnegie Mellon faculty and graduate students with investigators and students in Portugal will include, among other, looking at technical, economic, regulatory and risk-related studies of distributed energy systems, micro-grids, and advanced information technologies for integrating and controlling such systems in conventional electric power systems; research on advanced methods for the instrumentation and control of electric power transmission and distribution systems in the face of normal operations, and risks from natural disruptions and terrorist events; studies of technology innovation in energy and environmental technologies, its relationship to government actions (both "carrots" and "sticks") and its implications for policy and energy systems development.

Finally, this subarea will also incorporate research on “engineering and policy applied to the software industry”. Software and software-related industries are now considered to be critical industries by many developing economies, and inordinate amounts of resources are being expended to develop these industries. This proposal intends to address the sustainable development of the Portuguese software and software-related industries by understanding their current characteristics, both in what concerns the technological products and associated services. Understanding the seminal R&D; context that might have given origin to each novel Portuguese software company is also relevant. Furthermore, it is also important to understand the role of formal R& D activities undertaken in Portuguese research centers and universities dedicated to software engineering subjects. This characterization will allow an in-depth observation of the organizational capability of the Portuguese software industry. This study will promote, within the Portuguese software industry, the need to adopt an increasing involvement with innovation projects to seek better software of a higher quality, on time, in the best way and at a lower cost.

As a result of its multidisciplinary and research nature, this proposed collaboration will be linked to other initiatives within the Area of Technology, Management and Public Policy, in particular the proposed PhD effort in Technological Change and Innovation involving UCP and UTL as well as existing Ph.D. programs at CMU. Other potential partners in Portugal should be defined during the initial phase of the Program. At CMU, the Department of Engineering and Public Policy (EPP), the Heinz School and the School of Computer Science by means of their Software Industry Centre (SWIC) will be involved. Collaborations with the Institute for Software Research International (ISRI) and the Software Engineering Institute (SEI) are also envisioned. This program will strongly involve the Portuguese software companies, by promoting the systematic adoption of industrial-driven PhD dissertations in the context of R&D; initiatives. The establishment of a Portuguese Software and Systems Process Improvement Network (SPIN) and the creation of a non-profit centre to contribute to the development of the Portuguese software industry might be considered as a potential initiative of the Information and Communication Technology Institute (ICTI).

The main goal of this project is to develop advanced programs of research and education in specific areas of mathematics of common interest and expertise between faculty of CMU and from Portuguese Universities.

This area considers a doctoral program in Applied Mathematics involving a consortium proposed and to be launched with the three Mathematical Departments at the three public Universities in Lisbon, namely, Universidade Técnica, through IST, Universidade de Lisboa, through Faculdade de Ciências, and Universidade Nova de Lisboa, through Faculdade de Ciências e Tecnologia, and the Math Department at CMU. Other potential partners in Portugal will be defined during the initial phase of the Program.

The work program defined for this area will reinforce the scientific and educational programs at the three Portuguese Universities in Applied Mathematics, stimulate mobility and scientific interactions among graduate students, researchers, and post-docs, and attract strongly motivated students that are able to integrate advanced research in Mathematics and applications.

The project establishes a program of graduate studies in Mathematics. In particular, it sets up an internationally attractive graduate program leading to a dual degree in Mathematics and promotes matching between members of CMU and members of the Portuguese Universities in order to ensure good scientific interaction and top research work.

http://icti.math.cmu.edu/