Back Ten years of UPF's participation in the Industrial Doctorates Plan: a direct route to knowledge transfer

Ten years of UPF's participation in the Industrial Doctorates Plan: a direct route to knowledge transfer

Over its ten rounds of applications (2013-2022), the Plan, initiated and supported by the Generalitat de Catalunya, has provided funding for a total of 59 thesis projects linked to various different doctoral programmes at UPF. Of these,  23 have already finished, 22 are still active, while the others were unsuccessful.


Imatge inicial

2013 was the year UPF received the first grant to carry out a thesis project within the Industrial Doctorates Plan. This particular project was managed within the Law Doctorate programme, in collaboration with the Cuatrecasas Law Firm, Gonçalves Pereira. Since then, and after ten editions, there have been 59 UPF-linked projects which have been funded, seven of which were approved in the various different partial resolutions published in the 2022 call for applications.

The industrial doctorates initiative (in which the doctoral student develops his or her research training - the aim of a doctoral thesis - between the university and a company, institution or other entity), is open to public and private universities within the Catalan university system. It is promoted by the Secretary for Universities and Research of the Generalitat de Catalunya government and managed by the Agency for Management of University and Research Grants (AGAUR) in collaboration with the Catalan Foundation for Research and Innovation (FCRI).

Until the 2022 call for applications, the programme was intended to cover three years (with the possibility of extending it to four) and the people chosen would receive a minimum gross annual income of 22,000 euros

After a pilot scheme which went ahead in 2012, the following year the programme started to offer an open call for applications each year, with two pathways: co-financed or with specific grants, and with termly resolutions. In the majority of cases, the business and the university submit a joint research proposal which is first reviewed, and then published to make it accessible to potential candidates. The company and the university research group have the final word in choosing which student is selected for the industrial doctorate. Until the 2022 round of applications, the programme had covered three years (with the possibility of extending it to four in 2023) and the people chosen would receive a minimum gross annual income of 22,000 euros.

Testimonial of a doctoral student who liaises between DTIC-UPF, SIRIS Academic and his university tutor.

Nicolau Duran Silva, a UPF Computer Engineering graduate (2018) and Master in Language Technologies from UNED, has been doing an industrial doctorate since September 2022, working between the Large Scale Text Understanding Systems Laboratory of the Automatic Natural Language Processing Group (TALN) from the University's Department of Information and Communication Technologies (DTIC) and SIRIS Academic, a consultancy in the field of higher education and research, which belongs to a non-profit foundation.

To carry out his doctoral project (which was one of the seven successful projects linked to UPF doctorate programmes in the 2022 call for applications) he explains how he "depended on a grant from the Generalitat, which gave me two clear advantages: the existence of a framework, with its respective milestones, and the co-funding given to the research group. In this way, it fosters commitment, interest and support for my research while also strengthening the business-university link.

His doctorate fits within the field of natural language processing, an artificial intelligence (AI) specialism which seeks to understand human language. "In particular, I'm exploring some problems related to extracting information from scientific and technical documents, such as being able to classify them into complex thematic categories, identify organisations and simplify the content. The practical application of all this is to be able to provide evidence to support and assess the public policies for higher education and R&D", he says.

Nicolau Duran: "The university gives me more formal research methodologies, and a great deal of knowledge as well as exchanges with colleagues who have a depth of experience in this field"

Despite having some difficulties in focusing on some of his doctorate tasks ("it's always more difficult working in two places as opposed to just one"), and that the pace of work at the company and the university are quite different, he assures us that he is gaining positive insights from both environments: "The university provides more formal research methodologies, and a great deal of knowledge as well as exchanges with colleagues who have a depth of experience in this field". On the other hand, he explains that "the business environment gives me a more critical and interdisciplinary perspective, which is often easy to lose when you go into a lot of depth on a particular subject."

In comparison with a "traditional" doctorate, Nicolau Duran explains that the industrial doctorate offers students the opportunity to work on real problems, it has a much more practical component, and it helps to define research which can have a greater impact, more quickly. "I think the interaction in the business-university space, on a training level, is really interesting: on the one hand, I've got the academic stimulus, there is a training and development plan, which gives me the chance to come face to face with the research community, but at the same time I can see the direct contribution I'm making".

In the business environment, the teamwork is what particularly struck him, as he works shoulder to shoulder with people who come from very different disciplines. "I think the university still has to learn a lot about that. I don't think it makes sense to put a group of people together who are all experts in the same knowledge area: that's missing out on all the potential and the different perspectives that are generated within diverse groups", he reflects.

Horacio Saggion: "The industrial doctorate offers the possibility for our research to make contributions to society in a very specific way, through doctoral training, and that gives us a lot of satisfaction"

Horacio Saggion, Professor of Computing and Artificial Intelligence and head of the Large Scale Text Understanding Systems laboratory of the TALN group (DTIC-UPF) is Nicolau Duran's thesis project tutor at the University. According to Saggion, the industrial doctorate has a clear component of direct knowledge transfer and technology: "it brings students closer to a reality that they wouldn't otherwise be exposed to until the end of their studies' ', he says.

The DTIC professor believes that this two-way engagement between the university and businesses really works: "In the first place, businesses have specific problems and needs that we at the university can provide solutions for. These problems represent challenges not only for the company, but also for researchers: the connection is clear. Secondly, the industrial doctorate, through doctoral training, offers the chance for our research to make contributions to society in a specific way, which gives us a lot of satisfaction."

Nicolau Duran at SIRIS Academic

59 UPF-linked projects among the successful candidates from 2013 to 2022

Since the Industrial Doctorates Plan began with its first call for applications in 2013, and until 2022, UPF has received support for a total of 59 projects, linked to doctoral programmes in Information and Communication Technologies (24), Biomedicine (22), Law (5), Political and Social Sciences (3), Communication (3) and Translation and Language Sciences (2). The average number of UPF projects featuring in the annual call for applications has been around seven, in some cases rising to nine (2019) and in other years, three (2015, 2018).

Among these 59 funded projects linked to UPF (41 men and 19 women, representing a proportion of 69% to 31%) there are a variety of different cases and situations. Of these, 22 projects are currently active; 23 were completed within the timeframe; seven of the doctoral students withdrew before completing the project and in two cases they withdrew before starting. Finally, there are five projects (all within the Biomedicine doctorate, two of them from the 2022 call for applications) in which UPF only participates as a centre where the students are enrolled, despite the fact that University does not receive any specific funding.

Of the 59 projects, 49 were obtained via the co-funding option (which includes funding for the Research Group) and ten were obtained via the specific grants option (aimed at responding to the current state of international and global research). Since the 2016 call, depending on the type, the grants amount to around 8,500 euros (specific grants) and 34,000 euros (co-funding) which is used to cover a range of costs such as enrollment and student mobility.

Summary of the seven industrial doctorate projects linked to UPF from the 2022 call for applications

Developing methods of natural language processing in the fields of science, technology and innovation

"NLP for the classification of science, technology and innovation literature is the title of the project being implemented by Nicolau Duran, and guided by Horace Saggion at UPF, head of the Large Scale Text Understanding Systems laboratory from the Natural Language Processing Research Group (TALN) at the DTIC, together with the company SIRIS Academic, where he has Ruggero Cortini as his tutor.

Research in the field of Natural Language Processing has progressed a great deal in recent years. The incorporation of techniques based on deep learning and the appearance of the latest generation of pre-trained language models (such as BERT, CGT and their successors) have changed the rules of the game in just a few years. The application of these techniques has been especially relevant for documents of scientific literature or patents, and even in specific domains such as biomedicine or clinical medicine, where specific challenges can be found because of the level of complexity.

The aim of this industrial doctorate is to exploit and develop methods for natural language processing and automated learning for the mapping of science, technology and innovation activities in different domains, on heterogenous textual registers obtained from different repositories, research projects, patents, news or social media.

Real-time, virtual skin treatment simulation for skin problems

The "Real-time skin treatment simulation" project is being implemented by Marcelo Sánchez and directed by Coloma Ballester, coordinator of the Intelligent Multimodal Vision Analysis Research Group (the IMVA group, previously IPCV) from the DTIC-UPF, together with the company Crisalix Labs where he has Gil Triginer Garcés as his tutor.

The simulation of skin treatments, such as wrinkle or acne removal, has practical applications in the field of cosmetic and aesthetic medicine. These problems are often dealt with in two stages: first, the area to treat is detected using an image. After that, the area is replaced by simulated skin which has been treated. However, the application of these techniques is a real challenge in an augmented reality environment, where the algorithms have to be run in real-time on mobile devices.

The thesis will develop lightweight architectures which simultaneously tackle the segmentation and inpainting problems, and it will optimise them to allow for real-time inference on mobile devices. This research has already borne its first fruits, having been published in the Asian Conference on Computer Vision 2022 and has also been incorporated in the technology used in simulations of plastic surgery by Crisalix.

Molecular Neural Network Potentials for Drug Discovery

The thesis project "Molecular Neural Network Potentials for Drug Discovery" is being carried out by doctoral student Carles Navarro, directed by Gianni de Fabritiis (ICREA research lecturer, linked to UPF's Department of Medicine and Life Sciences and the Institute of Medical Investigation, IMIM, where he is head of the GRIB research group on Computational Science), together with the company Acellera Labs, where he has Majej Majewski as his tutor.

The research aims to develop new molecular potentials on different scales based on artificial neural networks and applied to the fields of structural biology, computational chemistry and drug discovery. The project will benefit from the active development of Neural Network Potentials (NNP) at Acellera and it will extend them to new areas of drug development. The doctoral student is participating in a training pipeline for novel NNPs and their application to real drug discovery problems; he is analysing structural data sets; exploring new neural network architecture perspectives, and testing and validating NNPs for their potential use.

Evaluating Labour inclusion policies from a gender and intersectionality perspective

The project "Evaluating Labour inclusion policies from a gender and intersectionality perspective: why do women lag behind in employment?", is being developed by Alberto Fabra, under the guidance of M. José González, lecturer from the Department of Political and Social Sciences (in which she is a member of the Sociodemographic Research Group, DEMOSOC, and the Centre for Gender Studies ,CEDGE), together with the Intermedia Foundation, where he has Sonia Moragrega Pallarols as his tutor.

The project is affiliated to Barcelona City Council, an institution which has two employment policy programmes. These programmes include a comprehensive route to employment from the moment someone starts the programme, offering them free training and education services in order to improve their employability.

The main aim of the research project is to analyse the factors which affect sectors of the population at risk of social exclusion from a gender and intersectionality perspective, in active employment policies, at all stages of the process and the differences in the results of these policies, as well as the different circumstances that eventually influence the final result.

MRNA-based vaccine designs and delivery systems to increase immunogenicity

The thesis project "mRNA-based vaccine designs and delivery systems to increase immunogenicity" is being developed by doctoral student Diana Wortmann, under the direction of Andreas Meyerhans, head of the Infection Biology Laboratory in the UPF Department of Medicine and Life Sciences, together with Hipra Scientific which provides the business context, where Antoni Prenafeta Amargós is her tutor.

During the SARS-CoV-2 pandemic it was demonstrated that vaccines based on mRNA could be produced very quickly and they are highly effective in inducing specific antibody and cell immunity. Nevertheless, they have some significant shortcomings. The current vaccines are only designed to provide an immune response through antibodies and they do not have immunogens specifically designed to activate T lymphocytes.

The aims of the project are, first of all, to develop simplified mRNA formulations that are immunogenic by alternative routes of administration; secondly, to develop a formulation that can be stored in simple refrigerated or unrefrigerated conditions in order for them to be more applicable for global use; and thirdly, to evaluate different immunogen constructs with respect to their capacity to induce virus-specific antibody and cytotoxic T cell immunity.

Developing a liquid biopsy test for the detection of early stage colorectal cancer

The project "Development of a liquid biopsy test for the detection of early stage colorectal cancer suitable for population screening" is being carried out by doctoral student Cristina Tuñi i Dominguez, under the tutelage of Roderic Guigó and the direction of Silvia Pérez-Lluch, researcher at the Computational Biology and Health Genomics Lab at the Centre for Genomic Regulation (CRG), in this academic research centre, and together with the company Flomics Biotech, with Julien Lagarde as director within the business context.

Colorectal cancer (CRC) is the third most prevalent form of cancer in both women and men. Despite current improvements in the treatment of this disease, CRCs still cause two million deaths in Europe every year. Almost half the CRC cases diagnosed today are detected at stages III or IV, when the chances of recovery fall to 64% and 8% respectively. Unfortunately, current screening methods, such as the immunochemical fecal test (IFT) are not ideal, and there is no alternative to the colonoscopy as a diagnostic test.

The aim of this project is to develop a cost-effective, non-invasive diagnostic method with greater sensitivity and specificity, by studying circulating long RNAs for the detection of cancer. She wants to do this through the use and development of computational methods and automated learning algorithms designed for population screening.

Improving efficiency in the initial phases of drug design

The thesis project "Evaluating the ligand-receptor affinity based on quantum-mechanical descriptors" is being developed by Brian Medel, directed by Jana Selent (linked to the UPF Department of Medicine and Life Sciences and the Hospital del Mar Medical Research Institute, IMIM, where she leads the GRIB research group GPCR Drug Discovery). It is being carried out within the academic context of the Medical Research Institute Foundation (IMIM Foundation) together with Pharmacelera which provides the business context, where Javier Vázquez Lozano is his tutor.

In the initial phases of drug design, when computational processes are used to find candidate molecules and turn them into medicine, simple and less precise chemical models are used in order to satisfy the competing demands of calculation time (cost) and precision of results: increasing this precision increases the chances of success for the candidate molecules in the later stages of the drug design.

This research project aims to extend the application of hydrophobic descriptors (molecules which do not chemically interact with water molecules) derived from quantum-mechanical calculations in the receptor-based molecular design. Therefore, the project deals with the design of new methodologies for applying these descriptors while also considering the structural information of the target biology.

*For these last two projects, where the students are enrolled on a biomedical doctorate programme at UPF, and are being carried out at the CRG and the IMIM Foundation respectively, Pompeu Fabra University does not receive any funding.

Reflections on UPF participation and some suggestions for improvement

Núria Sebastián Gallés is professor of Psychology and coordinator of the Speech Acquisition and Perception Research Group (SAP) from the Center for Brain and Cognition (CBC) within the DTIC, and has been director of the UPF's PhD School for almost ten years: from July 2013, when the school was set up (and coinciding with the first calls for applications for the industrial doctorate) until the beginning of March 2023, she has gained experience which gives her enough perspective to be able to reflect upon UPF's participation and point out some suggestions for improvement.

"If we bear in mind UPF's profile principally within the social sciences and humanities, I would say that the success of our participation has been moderate, especially if we compare it to the great success that the University has had in this field in terms of the number of projects obtaining funding from the European Research Council (ERC), which we can take as an indicator of the quality of research that an institution does. In the technical-scientific fields we have done quite well, but in the fields of social sciences and humanities, we could obviously do better", states Núria Sebastián, who hopes that there will be a new drive from the new governing team in this regard.

In terms of the Industrial Doctorates Plan, Núria Sebastián thinks that they should be more clearly defined, especially because, as she says on her website, "it will focus their objectives of contributing to competitivity and the internationalisation of the 'Catalan industry', but at the same time the doctorate can also be carried out within a company, institution or NGO, not necessarily in 'an industry', and this is a tension (or even a contradiction) which the programme has and which limits its implementation", she argues. She also points out that now they are celebrating ten years of the plan, the Generalitat should carry out an in-depth analysis to evaluate whether the programme is successful or not, and in particular to make a comparison between the industrial and "traditional" doctorates and their links to business and research.

Finally, Núria Sebastián explains that within the public sector (both public administrations as well as universities) it is clear that efforts are being made and there is interest in encouraging closer relationships between the business and university worlds, but that she does not see a corresponding interest from the business sector, contrary to what happens in the most dynamic contexts around the world, such as the USA or the Netherlands: "The business world doesn't invest in science as much as they should and, as such, with a relative interest in hiring highly qualified people, with specific training in planning, public speaking, teamwork, critical and creative thinking, with international experience, etc. well, these are the cross-disciplinary skills that are often acquired through a doctorate", she concludes.



SDG - Sustainable Development Goals:

Els ODS a la UPF