Portugal AI Strategy Report

In June 2019, the Portuguese Government presented the national strategy AI Portugal 2030 (Portugal, 2019) to set out challenges and opportunities of the growing AI ecosystem in Portugal. This strategy presents the plan to foster the use of AI in the public and private sector during the coming years. The plan concentrates its actions on inclusion, education, qualification, specialisation and research as people are the main engine of a successful AI deployment.

The national AI strategy of Portugal does not disclose financial figures, or estimations, for its implementation.

Portugal AI Policies on OECD.AI dashboard

One of the main strategic objectives is human development, particularly the enablement and reinforcement of the population vis-à-vis challenges and priorities set by the AI technologies. This objective will go along of its main axes: increasing the overall level of AI education, upgrading the labour force qualifications, and fostering the specialisation in AI fields.

The strategy emphases the importance of digital and AI education for the future generations. In this respect, a well-developed education infrastructure will include training in AI for earlier educational levels as much as for higher ones like bachelors, masters, post-graduates and PhDs. Finally, the strategy aims to encourage students' specialisation in computer, data science and STEM subjects to create a highly qualified labour force in AI related areas.

The following actions aim to nurture talent and skills in AI:

• Teaching young students fundamentals of machine learning through the Ciência Viva Clubs;
• In 2019, as part the programme Ciência Viva no Laboratório - Criar Futuro, 270 high school students have participated in a two-weeks summer school aimed at gaining literacy on AI applications and societal impacts, data science and machine learning;
• In 2019, as part of the 6ª Conferência Professores Espaciais, the European Space Education Resource Office (ESERO Portugal) along with Ciência Viva promoted a lecture on Data Science and AI and a workshop on AI applications for space education. A hundred teachers participated in the lecture and sixty teachers participated in the workshop;
• In 2020, the Director-General for Education and the SeguraNet Awareness Centre have launched a MOOC on AI in education. The first session counted about 2000 participants. The second session is ongoing;
• In the last two years, public higher education institutions have launched about twenty graduate and postgraduate degrees on data science and AI. Specific courses have also been introduced in ongoing “classical” postgraduate degrees;
• In 2020, higher education institutions have launched several executive programmes, advanced courses, summer or winter schools on AI-related (e.g. big data, machine learning, business analytics, computer vision).

The strategy targets the most skilled individuals as the labour force to enrol in AI and data science qualification programmes. Increasing the stock of active employees with digital and AI skills is crucial to meet the new labour demand in AI. The strategy supports actions to increase digital and science competencies in the public sector.

The following actions aim to re-skill and up-skill citizens and to offer lifelong learning opportunities:

• Creating Regional/local Networks for Digital Qualification to re-skill and up-skill professional training, e.g.: the Portuguese polytechnic institutes (located in different Portuguese regions) are offering short courses under the Upskill Programme. This programme involves different stakeholders (e.g. companies, higher education, public administration institutions) and aims at upskilling qualified people who are unemployed to meet the ICT-oriented labour market demands. The first edition covered training areas relevant to AI, such as Systems Applications and Products (SAP) – Analytics (e.g. competences on data processing for Business Intelligence and Data Analytics), and Customer Relationship Management (e.g. competences on supervised and unsupervised learning and automatic processing);
• Setting up e-learn platforms with courses on AI and other specific application domains: in 2019 the NAU – a massive open online course platform - developed by the National Foundation for Scientific Computation – was launched for the Portuguese-speaking communities. Higher education institutions, and research centres can disseminate their courses through this platform;
• In 2019, the National Institute of Administration (INA) has defined digital skills as one of its three new strategic areas. The Training Programme for Digital Transition in Public Administration – one of the four programs under the digital skills strategic area – aims to empower works with skills in the emergent digital technologies essential to public service (i.e. AI, big data and innovation in digital environment);
• Supporting HPC Competence Centers - national centers that offer education and training opportunities on HPC and related advanced computing tools.

From the lab to the market

Supporting AI products and services from the lab to the market will stimulate knowledge-intensive research and the entrepreneurial ecosystem on AI. Policy actions aim to create a community of young, vibrant, knowledge-intensive companies that can develop AI cutting-edge technologies. To this purpose, the following policy initiatives establish a strong research and entrepreneurial ecosystem in AI:

• Supporting application-oriented and fundamental research, e.g.: in 2020, an International Research Programme together with an Observatory were launched in Montesinho. Earth observation is one of the five strategic areas of the Observatory, which includes the use of satellite information integrated in advanced processing systems and AI;
• Developing a Centre of excellence for AI R&D;
• Promoting new and innovative solutions to simply the public administration (SIMPLEX programme). Since 2018, public funding has also been specifically allocated to AI research and innovation aiming at public services and processes optimisation (i.e through projects such as DSAIPA and SAMA);
• In 2020, under a call on advanced computing, launched by the Portuguese Foundation for Science and Technology (national funding agency), eighteen AI-related projects out of 129 approved projects had access to RNCA supercomputers and Cloud infrastructure;
• Launching Innovation funding programmes;
• Creating sandboxes and testing facilities;
• Refining the innovation vouchers;
• Reinforcing the national funding management and identifying KPI for investment evaluation.

In addition, the Portuguese strategy promotes research and innovation in specific scientific areas. Accordingly, the following priority sectors will be promoted to "living labs" for experimentations in AI:

• Urban transformation (sustainable cities);
• Sustainable energy networks;
• Biodiversity (green and blue economy);
• Autonomous driving;
• Cybersecurity;
• Quantum materials;
• Adaptive learning curricula for students.

Specialised AI services will also make use of:

• Natural Language Processing (for automatic translation and other automatable services);
• Real time AI (for securing business and financial transactions);
• AI for software development;
• AI for edge-computing.

Other policy initiatives not explicitly mentioned in the AI strategy report, may contribute to the creation of a vibrant enterprise ecosystem in Portugal, such as incentive systems for technological R&D in companies, Startup Portugal, Incubation vouchers, Industry 4.0 national strategy, and the Action plan for the digital transition among others.

Policy actions towards advanced use and deployment of AI in the public administration include among others, LabX – aiming to create a culture of experimentation and innovation in the public sector –, and InnoLabs to share good practices across the public administration.

There is a widespread consensus that AI will profoundly transform our world with powerful solutions to many societal and economic challenges. To fully leverage this transformation, a collaborative and networking approach is essential. The Portuguese strategy favours the establishment of partnerships between public and private institutions, and it advocates that joint undertakings should include European and international collaborations. Support actions to networking and collaborations include:

• Extending collaborative laboratories (CoLabs) and Digital Innovation Hubs (DIHs). A renewed national DIH network will shortly be announced;
• Increasing partnerships with Member States through collaborations on electronics and systems (Key Digital Technologies), high-performance computing (EuroHPC), Smart Networks and Services (5G and beyond), and the Quantum Technologies (H2020);
• Fostering long-term collaboration between academia and companies through framework contracts and data/technology sharing platforms;
• Increasing the participation of the Portuguese entities in the European Framework Programme, Horizon Europe and Digital Europe Programme;
• Participating to European Networks (e.g. Big Data Value Public-Private Partnership), European AI excellence centres and other European DIHs (e.g. DIH on cybersecurity at Leon or the DIH on IoT in Salamanca). Portugal also participates in the EuroCC project on high-performance computing, the National Network on Advanced Computing (RNCA) and joint AI-related projects funded under the FCT-UTAustin Programme on Advanced Computing.

The strategy also emphasises dissemination campaigns to promote Portuguese AI in and out the national borders. These campaigns are essential tools of digital inclusion as they inform the whole population on the benefits of AI technologies. International campaigns instead are crucial to attract international AI and ICT talents to study and work in Portugal. Hence, the following actions will sponsor the national and international attractiveness of AI:

• Promoting the attractiveness of Portugal for foreign talents, students, researchers and experienced professionals, while reducing border obstacles;
• Supporting national stakeholders’ participation in the AI4EU platform, the EU AI on-demand platform and ecosystem;
• In February 2021, AI Portugal 2030 launched Forum IA, an online platform to raise awareness on AI among citizens and relevant stakeholders of the public and private sector. This forum also aims to foster networking among the participants.

Legal, regulatory and ethical frameworks are essential to develop standards in AI as for transparency, accountability, liability and ownership. In this respect, the Portuguese strategy proposes the following actions:

• Creating an ethical committee to define guidelines for AI and Automation;
• In December 2020 the Agency for Administrative Modernisation (AMA) has launched a guide (in beta version) for ethical, transparent and responsible AI;
• Supporting the development of a legal and regulatory framework to identify liabilities when AI decision making;
• Supporting companies and regulators to find appropriate legal frameworks.

As proposed in the Action plan for the digital transition, general principles have been established for the creation and regulation of Free Zones for Technology (FZTs). FZTs are physical spaces to support the demonstration and testing of new technologies, through the creation of specific and adapted regulatory regimes. These regimes aim to promote a culture of experimentation in Portugal, including for AI-based solutions and robotics. The government is finalising the legislative process for the creation of FZTs.

The Portuguese strategy envisages the following actions to support the AI infrastructure:

• Creating the National Data Infrastructure, a centralised repository for administrative data. This action also provides guidelines about data sharing in the scientific community, likewise the Open Data Policy of the Portuguese Foundation for Science and Technology (FCT);
• Establishing supercomputing and quantum computing facilities thanks to the Advanced computing Portugal 2030 (Portugal, 2018) that defines the objectives of high-performance computing in Portugal;
• In 2021 the new Vision supercomputer was installed at the University of Evora. It is designed to enhance the application of AI methodologies to the various national strategic domains. With a maximum performance of 10 petaflops, this supercomputer allows ​​machine learning and deep learning techniques to be tested and optimised within the National Network of Advanced Computing;
• Infrastructures for piloting and testing will be created for several technologies, including AI and robotics;
• The Sines 4.0, Hyperscaler Data Centre and EllaLink submarine cable contribute to foster interoperability and connectivity between Europe and Latin America.

Climate and environment

The Portugal Space Agency, together with the Foundation for Science and Technology (FCT), the European Space Agency and Unbabel Labs, launched the AI Moonshot Challenge with a EUR 500,000 award for combining AI and space technologies to tackle the major sustainability issues. Specifically, this competition calls participants to find solutions for the detection, location and monitoring of maritime waste along the water column on a planetary scale. Teams will have access to AI resources, Earth Observation Data from satellites and supercomputing capacities.

The MIT Portugal Programme collaborates with several partners to the Mission Control for Earth Challenge that envisages accelerating positive change for Spaceship Earth’s ocean and land subsystems to address the urgent needs of a changing planet and climate. This collaboration aims to employ AI tools, science and technology for the ‘betterment of humanity’.

COVID-19 pandemic

The Portuguese National Funding Agency for Science, Research and Technology (FTC), within the general context of its supports to R&D activities in the field of Data Science and AI in Public Administration, has launched a new competition aimed at data processing within the current pandemic of the COVID-19. The objective is to support R&D projects with an emphasis on supporting citizens and health care services. Twelve projects were approved, of which six with access to RNCA resources.

Connect Robotics from Portugal operates with the support of the European Space Agency, and during the COVID-19 offered a service of urgent delivery by means of drones. Connect Robotics could deliver within minutes goods that would take much longer to reach quarantined or locked-down populations. The solution consists of autonomous drones and a web platform for operation, using Galileo satellite constellation in addition to GPS, Connect Robotics attains a higher precision in locating individuals.

The Portuguese AI strategy will undergo annual reviews and updates by the Ministry of Science, Technology and Higher Education and the Ministry of Economy and Digital Transition.

Portugal (2019). AI Portugal 2030: Portuguese national initiative on digital skills. Coordination Office of INCoDe2030. https://www.incode2030.gov.pt/sites/default/files/julho_incode_brochura.pdf

Portugal (2018). ACP.2030 Advanced Computing Portugal.2030 strategy. Coordination Office of INCoDe2030. https://www.incode2030.gov.pt/en/advanced-computing-portugal-2030

Portugal on OECD.AI observatory 

Read full report