The Flemish AI plan aims to prepare Flanders for the successful use of AI. Three main objectives are put forward:
- Strengthening demand-driven and challenge-oriented basic research on AI
- Boosting AI in all sectors of business and government
- Using AI for the benefit of society and citizens
Each of these objectives is based on a section of the Flemish AI plan.
Track 1: strengthening top AI strategic basic research in Flanders
Demonstrators are being developed at the intersection of four major research challenges with four focus domains and practical applications.
The four research challenges are:
- AI-driven data science: being able to make complex decisions by learning (data) and reasoning (knowledge)
- AI "at the edge": being energy efficient, working locally at a device
- Multi-agent cooperative AI: being able to perform complex tasks together
- Human-like AI: smoothly collaborating with humans
The application domains are:
- Health (precision medicine, decision-making systems (both clinical and hospital),
- industry4.0 (smart energy, smart machines, smart factories)
- government and citizens (personal data, job placement).
The demonstrators serve to facilitate faster technology absorption by industry. Proof of concepts based on real datasets and requirements should illustrate what a technology can achieve in real life situations. They can also provide inspiration for new research questions.
All five Flemish universities (KU Leuven, UGent, VUB, UAntwerpen, UHasselt) participate, all four Flemish strategic research centres (imec, VIB, vito, Flanders Make) as well as Sirris. A total of EUR 12 million per year is allocated for this purpose. Approximately 150 PhD students, 100 (senior) researchers and 90 professors are involved in the activities.
The research part serves to (i) expand the current capacity and expertise of the Flemish AI field, (ii) deepen it and (iii) further embed it in Flanders. The success is measured by a number of indicators, in particular the increase of the number of impactful publications, the number of PhD students and doctorates obtained, the amount of additional funding obtained from other sources (especially European), the datasets and software that are made available, the number of patents filed and granted, the number of spin-offs and start-ups, and the number of collaborations with other actors.
The evolution of these indicators is monitored annually, and if necessary, remediation will follow. In addition to these quantitative indicators, qualitative analyses are also made of the "proof of concepts" and demonstrators and an annual review is undertaken by an international panel.
The last indicator is defined as a key performance indicator where the goal is that 300 unique actors are reached after four years, which represents an increase of 50% since a baseline measurement set this number at 200. The target groups constituted by are what are called "innovators", "early adopters" and "early majority“ (as defined in Rogers' Innovation Theory).
Track 2: implementation in the Flemish business community
To stimulate the use of AI in all sectors, both among technology users and providers, support is provided through three main types of activities and related tools:
- awareness raising, informing, coaching and guidance
- collective knowledge build-up and dissemination by research organisations via projects
- knowledge build-up and implementation at companies supported by subsidies
In a first phase, VLAIO wishes to inspire, sensitise and inform as many companies as possible about the potential of AI and to encourage them to take further action in function of their own AI-use case. For this purpose, collective actions (workshops, seminars, master classes, ...) as well as limited, individual coaching and guidance are offered. These actions are carried out with the support of VLAIO by the partners from the VLAIO network, with in particular the service providers from the entrepreneurship contract and the Flemish universities of applied sciences in the context of Blikopener. The VOKA digihubs give an extra impulse to companies to discover the possibilities of AI and other digital technologies in function of their own AI use case in an easily accessible way.
At https://www.digitaletoekomst.be/nl/artificiele-intelligentie, a web platform was launched that provides companies with inspiring stories and recent developments around AI applications. It also informs them about ongoing projects and support opportunities in the context of the AI.
In the second group of activities, through programmes such as COOCK (Collective Research & Development and Collective Knowledge Dissemination/Transfer), TETRA and the living labs industry 4.0, companies are given the opportunity to acquire knowledge at an early stage under the impulse of a research organisation about the latest developments in AI that respond to the needs of and/or offer new market opportunities to a large group of companies.
Finally, VLAIO uses its regular business support tools (R&D support, SME growth subsidy, AI-ICON,...) to support technology providers and technology users in their own AI knowledge build-up on the one hand and the in-house implementation of AI on the other. This can be done entirely autonomously or in collaboration with other companies or knowledge institutions. Where useful and relevant, companies can also participate in projects in an international consortium with VLAIO support. Through the VLAIO business advisors, companies are directed to the appropriate instruments and partners from the VLAIO network who meet their specific needs.
Quantitative indicators are provided per instrument and mainly concern the number of companies reached via the different types of activities and the number of approved and supported AI projects. An annual AI barometer survey is used to map the impact of AI on the companies (competitiveness, productivity and competencies). This track accounts for an investment of 15 million euros per year.
Track 3: flanking policies aimed at raising awareness, training and ethical framework
Societally relevant issues concerning the broad application of AI are taken up in the third track. Approximately 5 million euros are provided annually for this purpose. Three major axes are distinguished:
- education, training and vocational training (EUR 3 million per year)
- ethical legal and societal aspects (750,000 euros per year)
- social awareness and involvement (about 1.1 million euros per year)
The education and training component is in its turn divided into three parts: for the highly skilled, for secondary schools and on the work floor. Each component represents an annual investment of 1 million euros.
A Flemish AI Academy (VAIA) is dedicated to gathering, offering and developing education and training materials and supporting trainers. The VAIA targets doctoral students who have not followed a computer science course and want to apply AI in their research domain (e.g. with an adapted introductory training). This is the "doctoral school" branch. These courses (e.g. bootcamps, summer schools) differ from the regular offer of the participating higher education institutions. In addition, there is also a "continuing education" branch that offers short or longer term training courses to higher educated people that mainly takes into account the needs of the business community. This targets professionals who want to apply AI in their work situation (e.g.AI for general practitioners). This branch should function in a self-sustaining manner after three years. Within the VAIA, all Flemish universities and colleges work together to produce a diverse offer, which must be complementary to the training initiatives on AI by the Flemish sector organisations and / or other innovation actors.
In order to improve digital skills in general and the use of AI in particular at the workplace, up to 1 million euros per year can be allocated. In addition, it is examined annually whether additional support on a project basis can also be provided by ESF Flanders. The Department of Work and Social Economy works together with VLAIO for this.
The secondary schools in Flanders are not forgotten either. VLAIO works together with the Department of Education and Training to support secondary schools in building expertise and producing learning materials on digitisation in general and AI.
The Knowledge Centre Data & Society (KDM) investigates ethical, legal and societal aspects of AI and offers all kinds of tools to use and/or develop AI and data-driven applications correctly. The centre not only draws up policy recommendations for the governance but also develops guidelines for companies to avoid undesirable effects (such as, for example. the systematic disadvantage of a specific group of the population) in the development and deployment of AI. Of course, the centre also deals with themes such as privacy and data protection. To this end, it regularly organise consultations and works together with business umbrella organisations. The centre also monitors the acceptance rate of/trust in AI by the citizens by an annual survey. The Knowledge Centre Data & Society receives 750,000 euros annually as a subsidy.
The third component, dissemination of information and awareness-raising among the general public, accounts for some 1.1 million euros per year and includes a range of projects and assignments of varying nature, duration and size. This allows to take advantage in a flexible way of opportunities that present themselves or needs that arise. This can range from a series of short TV broadcasts to initiatives on citizen science in which, after a co-creation process, a number of research questions about climate, mobility and health will be further addressed and developed. In addition, a monitoring system (AI barometer) is also being developed to measure every two years the impact of the AI program in its totality on the business community.
