Bridging the Digital Talent Gap: Towards Successful Industry-University Partnerships

Bridging the Digital Talent Gap: Towards Successful IndustryUniversity Partnerships Bridging the Digital Talent Gap: Towards Successful Industry-University Partnerships March 2020 Report prepared by Informatics Europe (Enrico Nardelli and Cristina Pereira) and the Rapporteurs of the parallel sessions in Artificial Intelligence, Cyber Security, and Software Engineering with the support of the European Commission, Directorate‑General for Communications Networks, Content and Technology. Published by: Informatics Europe Binzmühlestrasse 14 8050 Zurich, Switzerland Supported by: European Commission DG Communications Networks, Content and Technology Unit F.4: Digital Economy and Skills BU25 1/163 B-1049 Brussels, Belgium © Informatics Europe, 2020 Disclaimer The information and views set out in this report are those of the authors and do not necessarily reflect the official opinion of the European Commission. The European Commission does not guarantee the accuracy of the data included in this study. Neither the European Commission nor any person acting on the European Commission’s behalf may be held responsible for the use which may be made of the information contained therein. 1 Foreword The rapid pace of digital transformation is changing every aspect of our lives – and is also generating needs for new skills and knowledge in the workplace. As job profiles evolve, shortages and mismatches can result, but universities may be able to help. Looking at current data, we are concerned that the gap between the demand for advanced digital skills and the supply of suitable employees is becoming entrenched. There are currently around nine million people employed as ICT specialists in the EU – two million more than five years ago – and, based on market trends, we expect this figure to continue growing. But already today posts for ICT specialists in the EU are going unfilled. This growing structural shortage must be addressed, in order to preserve the innovation potential and future competitiveness of European businesses and jobs. Universities have a strong role to play, but it is not always easy for them to adapt their curricula to rapidly changing needs, while at the same time providing a solid foundation for their students. To consider how we can best address the situation, we decided to co-organise the Digital Talent Gap Workshop on October 2019 in Rome, where we brainstormed ways of boosting collaboration between academic institutions and Industry. The workshop brought together representatives of universities, businesses of all sizes, and students, and focus on Artificial Intelligence, Cyber Security and Software Engineering. Participants were encouraged to share their insights and experiences, work together to explore the issues at stake, and co-design concrete solutions for the way forward. The results of this fertile discussion can be found in this report, which will feed into the current debate on future policies, including programmes at European level like the Digital Europe programme. Fabrizia Benini, European Commission and Enrico Nardelli, Informatics Europe 2 Table of Contents Foreword ........................................................................................................................................................... 1 Executive Summary ........................................................................................................................................... 3 1 Introduction .................................................................................................................................................... 5 2 Artificial Intelligence ....................................................................................................................................... 6 2.1. Methodology and Process ...................................................................................................................... 6 Issue 1. Talent Competition ....................................................................................................................... 7 Issue 2. Curricula ....................................................................................................................................... 8 Issue 3. Platform ........................................................................................................................................ 9 Issue 4. Ethics and legal considerations .................................................................................................... 9 2.2. Conclusion ............................................................................................................................................ 11 3 Cyber Security ............................................................................................................................................... 12 3.1. Scene Setter .......................................................................................................................................... 12 3.2. Methodology and Process .................................................................................................................... 12 Issue 1. Demand and Supply of Cyber Security Talents and Skills ........................................................... 13 Issue 2. Upskilling the labour force ......................................................................................................... 15 Issue 3. Building a pipeline of talent ........................................................................................................ 15 3.3. Conclusion ............................................................................................................................................ 16 4 Software Engineering ................................................................................................................................... 17 4.1. Scene Setter .......................................................................................................................................... 17 4.2. Methodology and Process .................................................................................................................... 18 4.3. Identification of Challenges and Solution Ideas ................................................................................... 18 Iteration 1. Collecting challenges and clustering them ........................................................................... 18 Topic 1: the inside view of Software Engineering ................................................................................... 18 Topic 2: the outside view of Software Engineering ................................................................................. 19 Topic 3: the interrelation of Academia and Industry .............................................................................. 19 Iteration 2. Identification of solution ideas and recommendations ....................................................... 19 Topic 1: the inside view of Software Engineering ................................................................................... 19 Topic 2: the outside view of Software Engineering ................................................................................. 21 Topic 3: the interrelation of Academia and Industry .............................................................................. 22 4.4 Conclusion ............................................................................................................................................. 23 5 Overall Conclusions and Recommendations ................................................................................................ 24 3 Executive Summary This report presents the results of the discussions held during the workshop Bridging the Digital Talent Gap: Towards Successful Industry-University Partnerships organized by Informatics Europe and the European Commission’s Directorate-General for Communications Networks, Content and Technology (DG CONNECT) in Rome on October 2019. The main goal of the workshop was to bring together Academia, Industry and Policy for an open dialogue in the search of solutions for the widespread problem of the digital talent gap in Europe. Currently, the EU is facing a systemic gap in terms of digital competences, including the most advanced. Almost all Member States face shortages for digital experts in all areas or ICT. If not urgently addressed this digital talent gap will consist in a serious threat for the future economic development of Europe. Taking advantage of their broad networks in higher education and Industry, Informatics Europe and DG CONNECT joined forces and brought together a distinct group of people to brainstorm and propose solutions to this challenging problem in particular for the areas of Artificial Intelligence, Cyber Security and Software Engineering. Each area was the central theme of five-hour parallel sessions, where participants joined together to debate the main challenges and stakeholders, and propose solutions and recommendations. Recommendations for Artificial Intelligence: • Universities need to develop more specific master courses in the area. • Universities need to develop curricula that are accessible to students without a specialized Informatics background (to support the creation of basic expertise in Artificial Intelligence also for experts in other disciplines). • Universities need to include in the Artificial Intelligence curricula contents that specifically address ethical and societal issues. • The education system needs to be more productive, insisting more on the increase of the number of skilled graduates in Artificial Intelligence, also including PhDs, which seems a rather viable solution to give EU small companies better chances to recruit talented Artificial Intelligence experts. • More effective collaboration between Academia and Industry should exist at the stage of design and shaping of new curricula, also as part of “National strategies for Artificial Intelligence”, which must give priority to actions that support this collaboration. • The creation of an EU institute for AI, as a permanent structure, that can support the development of EU tools and platforms for research and development in AI. This institute would favour aggregation of researchers, support the dissemination of EU tools in third-level education and provide companies with a technically competitive basis to develop AI-based solutions. Recommendations for Cyber Security: • Create a common understanding about the set of knowledge and skills needed to qualify Cyber Security experts. • Explore new ways to strengthen collaboration and create permeability between businesses and university, dialogue and constant exchange, allowing experts from Industry to contribute to teaching and vice versa. This collaboration is crucial to improving the match between supply and demand of Cyber Security skills, but also to ensuring mutual learning, pooling resources and overcoming constraints. 4 • Enhance and expand the offer of Universities and other higher education institutions, which should adopt more flexible approaches and, for example, design flexible learning pathways that allow students to combine academic assignments with hands-on experience in businesses. • Extend the reach of Universities and other higher education institutions, which should provide more options, for example, for students from other disciplines who want to turn towards ICT related studies, drop-outs or professionals in search of a career change or upskilling opportunities. • Build a pipeline of Cyber Security talents by combining actions at different levels. Already in primary and secondary schools, it is necessary to strengthen the understanding of security risks posed by digital technologies and raise awareness of the main mechanisms to protect one's own digital devices and data. Recommendations for Software Engineering: • Improve the image of Software Engineering and a broad public understanding of its benefits. This might be achieved by a much closer cooperation between Academia and Industry. • Modern Software Engineering curricula should benefit from such a close collaboration. This might happen by dual study programs, where realistic industrial case studies are closely coupled with conceptual and fundamental aspects taught within the academic Software Engineering study program. As a base, convincing industrial examples showing the benefits of Software Engineering should be collected and should be made accessible in open-access repositories. • Modern Software Engineering curricula should comprise multidisciplinary projects where students learn to cooperate with stakeholders from possible application domains. • Certification programs for the Software Engineering profession as well as educational up-skilling programs in Industry should be developed. This will allow to clarify and improve the role and image of a Software Engineer in industrial projects for developing complex systems. • All these recommendations should be further elaborated and disseminated by a working group at European level where stakeholders from universities and companies jointly develop guidelines and recommendations on how to close the gap between a research-oriented education and industrial needs. Informatics Europe could bring its broad network of academic institutions and industrial research labs in support to initiate and lead such a working group. All parallel sessions have acknowledged that Europe needs a more open cooperation between Industry and University and that the European Commission and European associations in ICT have a key role in nurturing and fostering this cooperation. The challenges identified during the workshop equally affect all Member States, independently from their size and their industrial profile, and addressing them requires high levels of private and public investment. These challenges have to be tackled collectively, by pooling resources and expertise across Europe and encouraging Universities, research institutions and businesses to cooperate, increase the relevant training offer and contribute to create digital ecosystems able to attract, train and retain digital talent. The forthcoming Digital Europe Programme (DEP) will be a key instrument to trigger the digital transformation of the European economy and society. Among its actions, DEP will support the development and delivery of specialised Master’s programmes in advanced digital technologies, short term training courses and job placements, by bringing together Academia, Research and Industry. Joint action by all stakeholders across the European Union is crucial to increasing our talent pool and making sure that Europe and its Member States develop and progress together. 5 1 Introduction Technologies have always offered opportunities for competitiveness, innovation and societal improvements. Industrial machines, by empowering the physical strength of people have helped to increase productivity, delivering more products and better services with less effort and at lower costs. The quality of life has steadily improved in the past four centuries, due to this industrial revolution. Starting from the second half of the XX century a new breed of machines appeared, the “digital systems”. Largely immaterial, being based on software, they contain knowledge like books. However, differently from books, theirs is “actionable knowledge”, that can be put into action at our will 1 . They are therefore dramatically different from industrial machines, since they amplify the cognitive capabilities of humankind. Digital systems offer great opportunities for Europe to stay competitive, increase its potential for innovation and remain an inclusive society. Today’s most successful businesses are those that use digital technology not just to boost productivity and improve internal processes, but as a means of reinventing themselves: their operational models, their value chains and their customer relationships. Digital transformation is the term used: in order for it to materialise, people with proper scientific and engineering knowledge and skills to create, develop, roll-out, and use these digital systems are a necessary condition. This implies satisfying the growing demand for specialists in digital technologies, as well as up-skilling the EU's citizens and workforce in the context of rapid digitalisation. Currently, the EU is facing a systemic gap in terms of digital competences, including the most advanced. While digital experts are among the most demanded professionals, almost all Member States face shortages in key areas, such as software development and analyses as well as database and networks. 53% percent of companies who tried to recruit digital specialists report finding it difficult 2 . The limited availability of proper knowledge and skills is the most frequently cited obstacle to investment across the EU (expressed by 77% of companies) 3 . The demand for advanced digital competences has been rising strongly over the past decade, with an average growth in the number of ICT specialists in employment of almost 4% a year, over the last 10 years 4 . However, supply is not keeping up with demand and the EU already lacks around 1 million ICT specialists. An important element to consider in this scenario is the tension between the need of universities to provide solid foundations to their students, and the rapid change in competences demanded by business, deriving from the fast pace of technology. This situation is more acute in exponentially burgeoning fields such as Artificial Intelligence, Cyber Security and High-Performance Computing. In addition to that, the field has complex multifactorial ingrained problems that need to be addressed for Europe to reach a higher offer of highly skilled professionals. The most elementary being: (i) the low number of school pupils deciding for a higher education degree in Informatics 5 (resulting, among other factors, from the lack of proper teaching of the discipline in schools and the very low proportion of women among the incoming students) and (ii) the low graduation rates (a significant number of students enrol, but do not complete their studies) in Informatics Bachelor programs. Informatics Europe and DG CONNECT joined forces to organize a workshop 6 in Rome, in October 2019, intended to debate and propose concrete solutions on possible ways of increasing the specialised educational offer in the areas of Artificial Intelligence, Cyber Security and Software Engineering. The main goals were to discuss how cooperation among universities and business partners developing and deploying digital technologies could contribute to alleviate the problem of the digital talent gap. Each area was the topic of a parallel session lead by a facilitator and a rapporteur. In the following sections, we present the report with main findings and recommendations drafted by each session’s rapporteurs. 1 E. Nardelli. Informatics: the third “power revolution” and its consequences, April 2017. 2 European Commission, Digital Economy and Society Index (DESI) 2019, based on Eurostat data (isoc_ci_eu_en2). 3 European Investment Bank, EIB Investment Survey 2018. 4 Calculation based on Eurostat data (isoc_sks_itspt). 5 We use the term Informatics, which depending on the country or source can also be referred as Computer Science, Computing, IT, ICT, Computer Engineering). 6 https://www.informatics-europe.org/ecss/about/past-summits/ecss-2019/talent-gap-workshop.html 6 2 Artificial Intelligence Rapporteur: Daniele Nardi, "Sapienza" University of Rome, Italy Facilitator: Tanya Suarez, BluSpecs, Spain 2.1. Methodology and Process The session on Artificial Intelligence (AI) was arranged around the following four main specific issues: 1. Competition for talent from large multinationals could be preventing deep skills from being developed and inhibiting the growth of AI research groups and SMEs/Startups. 2. To respond to the rapidly changing nature of AI and the respective markets, more innovative curricula are required at third-level education. 3. Current skills and learning are focused on existing non-European technology platforms and stacks. 4. While the technical skills for AI are being developed, developers must also understand the ethical and legal implications of AI adoption. The methodology chosen to run the workshop, as suggested by the facilitators, was developed in two phases, that were preceded by a quick general introduction to the field (Daniele Nardi) and to the market needs (Tanya Suarez). After the introduction, five groups were formed spontaneously, with two groups focusing on the innovation of the curricula, given the number of interested attendees. Phase 1 aimed at identifying a set of ideas, through a prior analysis of Causes, Stakeholders and Similar Solutions. The first phase was held in the morning after the introduction and was concluded by a brief presentation of each of the five groups. Phase 2 aimed at developing one or two ideas emerged in the first phase, through the definition of Objectives, Impact, Key Phases, Resources. The second phase was held after the lunch break and its outcomes have been synthesized and presented to the general audience in the final session of the workshop by the rapporteur (Daniele Nardi). The program was undoubtedly extremely ambitious to provide well-articulated conclusions; however, the goal of reaching useful reflections and concrete suggestions for possible future actions has been to a large extent achieved. The outcomes of this session are summarized below, which includes four sections, one for each of the four above highlighted specific issues. For brevity, we will refer to them as: (1) Talent Competition, (2) Curriculum, (3) Platform, (4) Legal and Ethics. We close this chapter with some final remarks. 7 Issue 1. Talent Competition Competition for talent from large multinationals could be preventing deep skills from being developed and inhibiting the growth of AI research groups and SMEs/Startups. According to a study by PwC 7 , the potential global gain in GDP attributable to AI could reach $15.7 trillion by 2030. Much of this gain is likely to come from companies that did not exist three years ago. At present, Europe is home to 1,600 AI early stage Startups, with nine in ten focusing on the business-tobusiness (B2B) market, developing and selling solutions to other companies According to MMC ventures, in 2018 one in twelve Startups is an AI company, but only one of the top 10 global AI unicorns is European. In fact, it is from the UK 8 . Phase 1: Causes: the group in many ways pointed out the clear lack (or unbalance) of resources to be committed by European research groups and Startups in hiring talents in AI. Another relevant cause was identified in the unsatisfactory relationship between academic curricula and Industry, as internships and stronger collaborations may indeed be regarded as a missed opportunity at least for the SMEs/Startups. Finally, it was highlighted that the attractiveness of big companies is also due to the large data resources that they already own. Stakeholders: the identification of the stakeholders highlighted not only the expected triangle AcademyIndustry-Students, but also government, agencies, facilitators and ecosystems, proving that the issue is not confined to technical aspects, but it involves society at large. Similar solutions: very few similar solutions were identified, and they range from creating new specialists to specialized hiring formats. Ideas: the resulting ideas were largely focused on the creation of more curricula, with more funding to the third-level education system and the aim of improving the attractiveness of the field (not mentioned, but also relevant to gender issues). In addition, there were proposals for creating specific forms of collaboration between Industry and Academy, for example co-funding academic positions and involving mentors in collaborations Academy-Industry. Phase 2: In the second phase of the work, the group focused on the idea: “Increase the number of skilled people”. The objectives that are suggested for the development for this idea are: increase of numbers (students, staff, …) and improve the relationship with companies. The pursuit of these objectives is expected to bring an impact by enabling master and PhD students to work on «industrial» problems. In this way, students are more interested by the work in EU companies and, at the same time, companies get access to state-of-the art knowledge and technical solutions. The means to develop the proposed ideas are identified in the “National Strategies for AI”, which must give priority to the solution of the issues addressed here. As final remark, the group pointed out that bridging the gap in the talent competition is also a question of resources and the above proposed initiatives can reduce the gap, while really bridging it may require to mobilize a great amount of resources. 7 Sizing the prize. What’s the real value of AI for your business and how can you capitalise? PWC, 2017 8 The State of AI 2019: Divergence, MMC Ventures, 2019. 14 different targets and people with different backgrounds. Skilled lateral entrants, aged engineers, or talented previous drop-outs need to be engaged and considered. Challenge 1.2: a clear taxonomy of competences and mapping of the current offer is needed. Challenge: strictly related to the previous challenge, there is the need to understand what Cyber Security skills mean, which distinctive areas of skills exist and what is the current education offer. While considering the framework offered by the US National Institute of Standards and Technology (NIST), the Association of Computing Machinery (ACM) Body of Knowledge, ECSO WG5 and efforts by four pilot projects funded by the European Commission (ECHO, CONCORDIA, CyberSec4Europe, SPARTA), a harmonized taxonomy for Cyber Security jobs and skills is clearly necessary. Solution ideas: the European Union Agency for Cyber Security (ENISA) or the proposed Cyber Security Competence Centre and Network of National Coordination Centres (CCCN) may lead the definition of a harmonized framework, next to the mapping of the current education offer and specialised centres for Cyber Security. This endeavour could build on ongoing projects and initiatives at the EU level, such as: • The taxonomy developed by the Joint Research Centre in 2017 as part of the impact assessment of the regulation for the cyber competence centres. By end of 2019, a second version of the Cyber Security taxonomy will be published. • The Cyber Security education map prepared by ENISA (currently being updated), identifying the different universities and the type of Cyber Security programs and courses they offer. • The Cyber Security Atlas developed by the European Commission, identifying around 660 Centres in Europe. Challenge 1.3: lack of a comprehensive mapping of supply and demand of Cyber Security jobs, including at regional level. Challenge: skill needs and supply and demand for Cyber Security jobs differ from region to region, because of diverse economic specialisations. Therefore, a mapping of skill needs and demand-supply issues should be performed at regional and local level, for a more granular vision. Solution ideas: a European Observatory building on existing efforts in skills intelligence (e.g. by CEDEFOP) could lead a comprehensive mapping of both the demand and supply of Cyber Security jobs at regional and local level. For the mapping of the supply side, synergies with the European Commission Cyber Security Atlas could be sought. This initiative would help shape strategic discussions and design tailored strategies, involving all the relevant players in a given geographical area. Challenge 1.4: need for mainstreaming of Cyber Security skills in other areas. Challenge: there is a substantial number of jobs that are not “Cyber Security jobs” but require Cyber Security skills. Thus, Cyber Security programs combined with domain specific knowledge are needed, both for digital specialists and other professionals using digital systems in their work (e.g. secure code programs for developers, data privacy and protection programs for health care providers, etc.). Solution ideas: awareness should be raised as regards the need to include Cyber Security skills in programmes for jobs of a different nature. Hence, the solution proposed is to build programmes combining cross-cutting Cyber Security competences with specialized domain-specific ones, based on guidelines and incentives, for example by the European Commission. 15 Issue 2. Upskilling the labour force The education system has to provide students and the existing labour force with the skills and knowledge needed on the labour market. Challenge 2.1: lack of suitable resources, facilities, programmes Challenge: there are not enough suitable training resources, cyber labs, trainers and teacher training. Solution ideas: resources and facilities for training and learning need to combine an academic orientation and a setup relevant for businesses. A good example mentioned was the Fraunhofer Cyber Security lab in Germany. This would also allow more SMEs to be involved in Cyber Security programmes. The lack of teaching resources needs to be addressed by creating more permeability among higher education, businesses and the public sector. Professionals need to be incentivized to share their knowledge and experience with students, an action that would also contribute to make skills acquired at the University more relevant to businesses’ needs. Challenge 2.2: university leavers Challenge: often, students leave university with only basic knowledge, do not complete the studies or stop at the bachelor degree (because they are lured away by attractive jobs). The challenge is twofold: first, to create opportunities for university drop outs who have useful skills for companies; secondly, to create incentives for students to continue studying and get a master’s degree whilst working. Solution ideas: special courses could be created for drop-outs and career shifters. To keep students after a bachelor degree, there is a need for more “dual degree” flexible MSc courses, combining part time work in the Industry and study at the University, modelled along existing good practices in a number of Member States. This combination would help student to implement their Cyber Security skills and to better understand real-life security issues. In parallel, by offering more flexibility, these hybrid courses could contribute to retain students in education. Challenge 2.3: SMEs unaware of or unable to tap into training Challenge: most SMEs have a local reach and do not have information about or the means to participate in EU programmes devoted to SMEs. Solution ideas: in order to support SMEs, it is necessary to create an appropriate collaboration with SME associations, and promote networking through Cyber Security clusters and specialised regional innovation hubs. Issue 3. Building a pipeline of talent Building a pipeline of talent means attracting enough students in this field. These efforts need to start in school and need to be supported by all stakeholders. Challenge 3.1: attracting students Challenge: a major issue is to attract enough students to university courses in Cyber Security, including women. However, next to the lack of clarity on the skills needed in order to qualify as a Cyber Security expert, there is also a low awareness about cyber risks and the need to be protected. Solution ideas: building awareness of the importance of Cyber Security, and the implications of cyber risks, is a first step towards the creation of interest in this field and the attraction of students. Communication and awareness raising should start with pupils in primary school, and be continued in secondary schools, to 16 increase interest in these technical topics and facilitate the selection of a degree in Informatics or Cyber Security later on. Secondly, it is important to communicate about the job opportunities offered by an education in Cyber Security, the quality of these jobs, the possibilities to expand in other areas and the overall societal implications. Young people are motivated by the challenge of making the world a better place. Convincing them that digital technology is an essential tool to achieve this will attract more of them to enter into these degrees and professions. Student associations may play a major role in this respect. Challenge 3.2: transfer of knowledge Challenge: raising awareness about Informatics as a discipline at school is valuable, but there needs to be collaboration between higher education, companies and schools that, currently, rarely exists. Solution ideas: to address this issue, it is necessary to empower students to build a sort of cascade system, with the support of stakeholders (businesses, cyber law enforcement agencies, higher education). The idea is to create a cascade of trust to transfer Cyber Security knowledge and experience and raise interest. Students are the key connectors in this chain, so that knowledge is transferred from university to secondary school pupils, and down to primary school. Teachers and parents should get involved along this chain with an active role on Cyber Security awareness raising. 3.3. Conclusion The workshop session represented an opportunity to discuss a number of issues and provided several suggestions that can be further developed in the context of EU policies. At this stage, some general reflections can be made. First, there is a need to build a common understanding of the set of knowledge and skills needed for a person to be qualified as a Cyber Security expert. Such effort could build on and bring together the work already ongoing at the EU and international level. Secondly, exploring new ways to strengthen collaboration and creating permeability between businesses and university emerged as an important underlying topic. Dialogue and constant exchange between education institutions and businesses are crucial to improving the match between supply and demand of Cyber Security skills, but also to ensuring mutual learning, pooling resources and overcoming constraints. In this view, new models are required, allowing experts from Industry to contribute to teaching and vice versa, students and academic staff to develop experience in Industry. Strictly linked to the point above, another message emerged between the lines: Universities and other higher education institutions are called to adopt more flexible approaches and expand their offer, in order to remain relevant and respond to a rapidly changing context and Industry needs. On the one hand, this means designing flexible learning pathways, allowing students to combine academic assignments with hands-on experience in businesses. On the other hand, higher education institutions should provide more options and extend their reach to, for example, students from other disciplines who want to turn towards Informatics studies, drop-outs or professionals in search of a career change or upskilling opportunities. Finally, although the focus of the workshop was on higher education, a pipeline of Cyber Security talents can only be built by a combination of actions at different levels. In other words, already in primary and secondary schools, it is necessary to strengthen the understanding of security risks posed by digital technologies and raise awareness of the main mechanisms to protect one's own digital devices and data. 17 4 Software Engineering Rapporteur: Gregor Engels, University of Paderborn, Germany Facilitator: Alessandro Bogliolo, University of Urbino, Italy 4.1. Scene Setter Software is a key factor of any innovation of today. Advances in mobile data networks and embedded systems have enabled the advent of the software-everywhere era, where computational resources are available in cloud, as well as in mobile and IoT devices. Thus, the development of new software as well as the migration, evolution and adaptation of existing software, due to changed requirements, is at the core of new products or services in any application domain. This is paired with a highly increased complexity of nowadays software as well as steadily decreased time-to-market demands. Also, the role of humans as users of software has changed, as users are nowadays more demanding and expect systems that are easy to use and understandable. Thus, it is expected that systems adapt themselves (both in their user interface and behavior) to the needs of individuals to gain appropriate acceptance. Lastly, software has not only to comply with standard quality characteristics, but also to obey social, legal, moral and ethical values. All this has not only lead to an enormous demand for software engineers during the last decade, but in particular for highly-skilled people who are able to cope with the novel challenges of human-centric, intelligent software systems deploying novel technologies, while requirements and contexts are changing continuously. Modern software engineers need to have multifaceted skills. They have to understand application domains and be able to speak to domain experts with an education in another discipline. They have to master agile development techniques, to deploy novel technologies and tools, to incorporate algorithmic services based on AI and machine learning techniques. They have to understand the role of humans as users and prosumers, they have to understand the potential of new network models and connected devices, to be able to exploit frameworks to enhance productivity, and to cope with the right choice of values. There have been various international initiatives in the past, to come to a uniform view on what Software Engineering is and what is an appropriate curriculum for an academic degree program in Software Engineering. The most prominent initiatives are, first, the SWEBOK, the Guide to the Software Engineering Body of Knowledge (SWEBOK Guide). It describes generally accepted knowledge about Software Engineering. Its 15 knowledge areas summarize basic concepts and include a reference list pointing to more detailed information. The SWEBOK Guide has also gained international recognition as ISO Technical Report 19759. Second, the IEEE/ACM 2014 Curriculum Guidelines provide detailed information on an appropriate curriculum content for undergraduate degree programs in Software Engineering. While these documents provide excellent information about Software Engineering as a subject, as well as appropriate curricula, it is a matter of fact that there is a huge shortage of software engineers in Industry. And due to the very low numbers of Informatics students in academic studies, this shortage will tend to increase in the future. All this will seriously hinder a surely desired socio-economic growth in Europe. Thus, the main questions addressed during the parallel session were: 1. What are the required skills of modern software engineers in Industry? 2. What can we do to increase the number of modern software engineers in Europe? 18 3. What can we do to provide on-the-job learning and qualification processes to software engineers in Industry? 4. What can we do to attract more students to become a modern software engineer? 5. How can we incorporate Software Engineering concepts in studies of other disciplines? 4.2. Methodology and Process These questions were discussed by around 30 representatives from Industry, Academia, and Policy at the Software Engineering parallel session of the Talent Gap Workshop in Rome. The session was organized as a highly interactive event. It consisted of two iterations of 125 and 85 minutes, respectively. The first iteration started with a plenary brainstorming on exploring challenges and discussing ideas for reducing the size of the talent gap. Each participant contributed with several post-its naming possible challenges. At the end, the mentioned challenges were clustered by identifying three main thematic topics: 1. The inside view of Software Engineering 2. The outside view of Software Engineering 3. The interrelation of Academia and Industry Following, in the second iteration, the whole group was split into three subgroups according to the three identified thematic topics. Each of them focused on and refined one thematic topic and came up with identification of challenges, derived solution ideas and recommendations which can be implemented in short-term. The whole process was guided by the facilitator of the parallel session, while the rapporteur took notes and gathered the results. 4.3. Identification of Challenges and Solution Ideas Iteration 1. Collecting challenges and clustering them The outcome of the first iteration was a clustering of challenges named by participants into three thematic topics. For each thematic topic, we list some of the mentioned challenges. Topic 1: the inside view of Software Engineering The inside view of Software Engineering deals with a thorough understanding of what Software Engineering should cover thematically. This should directly be reflected in the design of appropriate Software Engineering curricula. While there are standardized descriptions of what Software Engineering is (e.g. SWEBOK) and how a standard curriculum should look like (e.g. ACM/IEEE Guidelines), it became obvious in the discussion that those documents were not well-known by the participants. Thus, the outcome was indeed a broad list of personal opinions about what Software Engineering is. Besides standard topics, the role of human factors, soft and communication skills, as well as social aspects and creativity in Software Engineering were stressed. There was also an agreement that there is a cultural shift from programming to model-driven development. Concerning learning and teaching Software Engineering, it was agreed that there is a need to teach more effectively, to let the students “feel” the need of Software Engineering techniques in commercial software 19 development. Life-long learning and mass education, as well as new forms of agile teaching, are key in Software Engineering. Finally, it was agreed that, to attract more talent, the fact that Software Engineering provides a great opportunity to solve and contribute to the 17 sustainable development goals defined by the UN should be mentioned to young people with a possible interest in the domain. Topic 2: the outside view of Software Engineering The outside view of Software Engineering deals with the image of Software Engineering as it is understood by stakeholders in Academia and Industry. A realistic understanding of what Software Engineering is, might help to make the profession of software development more attractive, as well as to prevent misconceptions of prospective students looking and deciding for the right study program. The outcome was, also in this case, a broad list of suggestions of what might help to improve the image of Software Engineering. Important suggestions were e.g. to clearly define what Software Engineering is, to clarify and valorize the role of a software engineer, and to understand how we can offer a much better perception of what we are as software engineers. This might even end-up in a new attractive name for Software Engineering that encompasses much better its close links to many other disciplines, and its social and creative aspects. Topic 3: the interrelation of Academia and Industry Besides an academic view on Software Engineering and an academic understanding of what are the right topics of a Software Engineering curriculum, it is important to clarify the needs of companies to define the right skills of software engineers. The outcome of the discussion was that a closer cooperation between universities and potential employers in the co-design of modern Software Engineering curricula might help. This might end up in a Software Engineering skills catalogue from an industrial viewpoint, which expresses that companies need problem solvers and application engineers with Software Engineering skills. In addition, university degrees might be coupled with university-led certifications, and a Software Engineering certification scheme acknowledged by the Industry might be set up. Iteration 2. Identification of solution ideas and recommendations In the second iteration, the three subgroups focused on and refined the thematic topics to identify challenges, stakeholders, causes, solution ideas, recommendations and needed resources, which might be installed in the short term. Topic 1: the inside view of Software Engineering Challenges: the main challenge is related to gaining an understanding of Software Engineering teaching at all levels of a University study programs, as well as in the context of re-training and upskilling of professionals. In addition, it was stressed again that there is need for a better understanding on how to attract more students aiming at becoming a software engineer, in particular more female students. All this 20 should be reflected in the design of appropriate Software Engineering curricula and training programs for these diverse groups. Stakeholders: Designing appropriate Software Engineering curricula and training programs involves many different stakeholders. They range from professors and teaching staff members to students on the university side, but also from ICT professionals to potential users of ICT systems on the industrial side. Causes: The participants agreed that there seems to be a strong misconception, by many stakeholders, of what a software engineer is. It seems not to be broadly known that there are different roles in Software Engineering, ranging from requirements engineers, architects, designers, coders to maintenance people as well as to project managers, method engineers, masters of agile methods and so on. An understandable definition of the mission of a software engineer is missing, something that young people can think of and decide to pursue, which includes also social skills and values. Solution Ideas: There was a strong agreement that already in the design of appropriate Software Engineering curricula and training programs all potential interested groups should be involved. It has to be conveyed that Software Engineering is about creating trustworthy, dependable software systems to support all aspects of human life and society. This does not mean programming and coding only, but also understanding the problem domain and define the boundaries between this domain and the software system, to identify and evaluate various different alternative solutions, to plan for and develop a solution, and to verify and certify it. Nowadays systems are complex systems, which have to be developed in a multidisciplinary way. Thus, students’ participation in multidisciplinary projects – across faculties at universities – has to be fostered. There should also exist a much better story-telling of what Software Engineering is. This might be done by a kind of Software Engineering manifesto, which includes examples from AI/Machine Learning and Informatics. Lastly, it should be shown more explicitly how Software Engineering can contribute to the 17 sustainable development goals defined by the UN. Recommendations: modern Software Engineering curricula should benefit from a closer collaboration between universities and companies. This might happen by dual study programs, where students focus on conceptual and theoretical aspects at universities and on practical ones at companies. The study content at universities could be improved e.g. by videos or webinars provided by industrial partners where realistic, industrial requirements analysis activities are shown to the students. Also, companies might open internal project management meetings to students as visitors to illustrate realistic case studies. On the other side, guidelines should be developed for universities on how to involve students in multidisciplinary projects during their Informatics study. On the industrial side, certification programs for the profession of a software engineer should be developed. This will allow to clarify and improve the role and image of a software engineer in industrial projects for developing complex systems. All these recommendations should be further elaborated by a working group at the European level where stakeholders from universities and companies jointly develop guidelines and recommendations on how to close the gap between a research-oriented education and industrial needs. 21 The results of the working group should be disseminated by an appropriate communication campaign in a broad way at all European universities and companies who deal with the development of complex software systems of the future. Resources: the intended working group as well as the planned communication campaign needs additional financial support to bring together experts, to conduct appropriate empirical studies, to develop dedicated guidelines, and to disseminate them. To support these actions, additional resources are needed. Informatics Europe could bring its broad network of academic institutions and industrial research labs in support to initiate and lead such a working group and to disseminate its results. Topic 2: the outside view of Software Engineering Challenges: the outside view of Software Engineering deals with a redefinition of its image in and for the society. As said above there seems to be strong misconception by many stakeholders of what Software Engineering is. This hinders the attraction of more software engineers as students and as future employees. In addition, this issue does not incentivize existing programmers in Industry to participate in upskilling programs to become a modern software engineer. Stakeholders: a better understanding of the benefits of high-quality Software Engineering affects many different stakeholders. This ranges from customers and potential future users of software systems to software developers and programmers on the industrial side, but also to researchers and teaching staff members on the university side. Causes: the benefits of applying rigorous Software Engineering techniques are at a first glance intangible. Thus, customers are quite often not willing to pay a huge amount for deploying Software Engineering techniques as they do not see any immediate impact on the quality of software. They do not understand that intensive efforts in e.g. eliciting requirements, in designing extensible software architectures, or conducting an extensive testing have an impact on software development costs on the longer run. In general, an understanding of the economical reward for Software Engineering is missing. Solution Ideas: Software Engineering as a discipline has grown up to be well established within the last 50 years. However, society still lacks an appropriate understanding of what Software Engineering is and what software engineers are doing. A communication campaign might enhance this understanding, by presenting real-life examples and demonstrating the practical impact of Software Engineering on our society and the life of individuals. Also, the image of the profession should be strengthened. This might be realized by introducing internationally recognized Software Engineering certifications for professionals which allows to check and state dedicated skills of a Software Engineer. Synergies can be exploited with existing awareness-raising campaigns aimed at bridging digital skill gaps. Recommendations: measures are needed to communicate the role and benefits of deploying Software Engineering techniques to a broad audience in society. Communication experts should be engaged to disseminate the characteristics of Software Engineering. Modern dissemination techniques like blogs, webinars and maybe even a dedicated TV soap might be used. Also, existing software engineers in Industry might be trained as ambassadors. As a base, convincing industrial examples showing the benefits of Software Engineering should be collected and should be made accessible in open-access repositories. Altogether, the discussion on the image and definition of Software Engineering should be renewed. 22 Resources: additional resources are needed for the planned communication campaign, for building the open-access repository of best practices and for a working group for renewing the definition of modern Software Engineering. Successful communication/awareness-raising/literacy campaigns can be taken as a model and possibly leveraged to expand the outreach and the impact. Informatics Europe could offer an excellent environment to initiate such a communication campaign, to set-up a repository of best practices and to lead the mentioned working group and to disseminate its results. Topic 3: the interrelation of Academia and Industry Challenges: Software Engineering is an engineering discipline with close links between research-based concepts and techniques and their applications in an industrial context. It is a discipline where both sides – Academia and Industry – can benefit of each other, and where each side needs the other one to improve. Research relies on concrete experiences with applying methods and tools in an industrial context, while Industry deploys research outcomes to improve efficiency and effectiveness of software development. Measures are needed for a better exploitation of research results in Industry as well as a better use of industrial experiences in an academic Software Engineering education. Causes: the amount of industrial software development has dramatically increased during the last decades. Thus, due to the insufficient availability of trained professional software developers, career changers were employed to work on software systems. In addition, due to the steadily increasing time-to-market pressure, software development focus quite often on a fast implementation by coding in a programming language, and spending less time on activities like a deep requirements elicitation or sufficient testing. Thus, there was insufficient capacity and skills available on the industrial side to disseminate successfully research results in Software Engineering to industrial applications. On the other side, advanced industrial ICT companies developed their own dedicated Software Engineering techniques as academic researchers did not always address the real needs of industrial software development. Altogether, the gap between academic Software Engineering research and industrial usage of Software Engineering techniques increased during the last decades. Solution Ideas: in order to make Software Engineering research results deployable in industrial applications and in order to let Software Engineering research understand the real needs of industrial software development, the gap between these two sides has to be decreased. Flexible academic study programs are needed that react to the dynamism of industrial Software Engineering advances in a timely manner. This might be realized by promoting dual programs where companies and university closely cooperate. In order to upskill industrial software developers including career changers, specific programs and measures should be established. Recommendations: specific projects on cooperative programs must be initiated, where both universities and companies have to be active players. Those projects might range from cooperative projects where experiences and best practices of Software Engineering are mutually exchanged to educational up-skilling programs. Mechanisms might be found to ease the participation of professionals in academic programs. This ranges from academic course materials which are made available to those who are already working in Industry to involving industrial professionals in academic teaching programs. Resources: additional funding is needed to realize above mentioned measures. As this is of high relevance for both Industry and Academia, a combination of public and private funding might be aimed at, especially for those activities involving up-skilling programs. Informatics Europe could bring its broad network of academic institutions and industrial research labs in support to initiate and coordinate those cooperative projects where Academia and Industry are involved. 23 4.4 Conclusion As software is everywhere and software is the key of any innovation nowadays, effective and efficient Software Engineering techniques are urgently needed for a healthy grow of the European economy and society. The discussions within this parallel session on Software Engineering have shown that there is a strong need in increasing the number of high-quality software engineers in Industry. Discussed measures are that the number of enrolled students as well as graduates in Software Engineering have to be dramatically increased. Also, the number of upskilled employees in Industry has to be increased. Furthermore, the image of Software Engineering and a broad public understanding of its benefits have to be improved. All this can only be achieved by a much closer cooperation between Academia and Industry, as modern Software Engineering techniques rely on the interplay of academic research results and industrial deployment and experiences. To achieve this, additional resources are needed for cooperative measures. Informatics Europe could offer an excellent environment to initiate and coordinate those cooperative measures on Software Engineering topics where Academia and Industry are involved.