Q&A with: Eindhoven AI Systems Institute (EAISI)

Q&A with: Eindhoven AI Systems Institute (EAISI)

This content was created by the Data Sharing Coalition, one of the founding partners of the CoE-DSC.

The Data Sharing Coalition supports organisations with realising use cases at scale to exploit value potential from data sharing and helps organisations to create required trust mechanisms to share data trusted and secure. In our blog section ‘Q&A with’, you learn more about our participants, their thoughts, vision, and ideas about data sharing. Daniel Kapitan works at the Eindhoven AI Systems Institute (EAISI). “As an EAISI Fellow, my role is to bridge the gap between the daily practice at companies and the applied research that we do here at the university. We need to strengthen the link between the academic world and the industry. I teach and I also support researchers to apply data and machine learning in their own research.” Daniel shares his thoughts.

1. Could you introduce your organisation?

Eindhoven University of Technology is a young university, founded in 1956 by the manufacturing industry, local government and academia. We educate students and advance knowledge in science & technology for the benefit of humanity. The Eindhoven AI Systems Institute (EAISI) applies the traditional strength of the university in high-tech systems engineering and our close ties with the regional industry in the Brainport region, to leverage the huge potential of AI in real-world applications in industrial engineering systems.

The university invested 100 M€ to establish the EAISI to ensure its top researchers from various research groups can work together in a newly established AI Laboratory to create new and exciting AI applications with a direct impact on the real world. EAISI has defined three main focus areas: medical technology, mobility and high-tech systems.

Issues related to trust and fair allocation of gains stand in the way of implementing data sharing platforms. I really believe that if we can offer more safeguards, we can win people over. All in all, it is a matter of trust.

2. To what extent is your organisation involved in data sharing (within and across sectors)?

Data sharing is an essential part of the research activities at EAISI. For example, within the health domain, we are preparing a Health Data Portal (HDP). The HDP will allow the sharing of medical data from diverse types of healthcare institutions in a secure and anonymous manner. It is a large-scale research collaboration between the Catharina Hospital, the Maxima Medical Center, Kempenhaeghe Epilepsy and Sleep Center, Eindhoven University of Technology and Royal Philips Eindhoven in the domains of cardiovascular medicine, perinatal medicine and sleep medicine.

In the mobility domain, the Responsible Mobility programme aims to improve the environmental footprint of mobility to achieve zero emission and zero casualty transport; which significantly reduces societal cost. To realise this, it is necessary to transform current transportation networks into smart systems that allow actors to make more informed and coordinated decisions while sharing tasks and resources. As an example, take the well-known problem of freight carriers. Even if you have an optimal planning, you only make optimal use of 50% of the trips, because you often drive back with an empty truck. If carrier A knows what carrier B’s schedule looks like, they can coordinate this, share tasks and resources and make the trips more efficient. To identify fair allocation (how much will carrier A and B get paid?) and align incentives between actors (e.g. a law that states we should have less empty carriers on the road) while solving complex transportation problems, the research combines cooperative game theory with data-driven algorithms and machine learning (ML) techniques that help discover underlying patterns. For example, what times are the busiest in public transport. By combining data we can gain new insights.

To empower the High Tech industry to become more socially, environmentally and economically sustainable for everyone, we need to create systems and factories that deliver more throughput, precision, intelligence and interoperability, year over year, while still reducing their negative impact on our world. In collaboration with industry partners, we realise innovative and disruptive solutions. Merging AI and engineering disciplines, we will pave the way towards designing, servicing and operating future autonomous, zero-waste high-tech systems and factories.

Besides the domain specific research areas, various groups within the TU/e investigate generic data sharing related subjects. The Coding Theory & Cryptology group focuses on privacy enhancing technologies including multiparty computation (MPC)* to enable secure data sharing. The Philosophy & Ethics group connects philosophy and ethics to technology, innovation and explores new frameworks to create data spaces.

3. Why is or should sharing data be important for your industry or domain?

As a research institute, one of our key goals is to foster open research that can be replicated. Data and data sharing are necessary to conduct relevant research across many domains, which is why we aim to make a tangible contribution to trustworthy data sharing and AI development. This is more easily said than done: as an institute, we are frequently challenged by issues related to data sharing, be it from a business, technological or governance perspective. Issues related to trust and fair allocation of gains stand in the way of implementing data sharing platforms. I really believe that if we can offer more safeguards, we can win people over. All in all, it is a matter of trust. In my work as a consultant, I mostly encounter data sharing issues. But it’s mainly a people problem, not a technological one. Fortunately, more and more people dare to discuss his issue.

4. What are the most promising data sharing developments and trends you see in your sector?

We believe that a combination of various trends will enable more large-scale data sharing and the creation of data spaces within the EU. Firstly, privacy enhancing technologies such as data synthetisation, federated learning and secure multiparty computation will be able to remove current roadblocks associated with security and privacy. Secondly, we see that standardisation efforts in many domains are improving interoperability of data. In the area of healthcare, for example, the adoption of the Fast Healthcare Interoperability Resources (FHIR) standard is encouraging and an important enabler for creating a health data space. Thirdly, we see that the AI field itself is moving away from a model-centric approach to a data-centric approach. Given the enormous progress that we have seen on the side of algorithms, it is now time to focus more on systematically engineering the data needed to build a successful AI system. In this light, researchers at EAISI call for a more prominent role of the Knowledge Scientist. The Knowledge Scientist is the person who builds bridges between data and business requirements, questions and needs. Their goal is to document knowledge by gathering information from business users, data scientists, data engineers and their environment to deliver reliable data that can then be used effectively in a data-driven organisation.

5. How do you see the future of data sharing, and what steps are you currently taking in that direction?

EAISI is involved in various ongoing projects to design and implement data sharing platforms. In the short term, i.e. within 1 to 2 years’ time, we aim to support organisations by creating more awareness, providing the necessary know-how and participating in consortia – like the Data Sharing Coalition. For example, we are involved in setting up a new initiative by Dutch Hospital Data (DHD), the Netherlands Comprehensive Cancer Organisation (IKNL) and Expertisecentrum Zorgalgoritmen, where both data and algorithms are shared across hospitals. Our role focuses on technical design and engineering, as well as providing a continuous influx of students who often take an exploratory, leading role in new projects and experiments.

For the long-term, we strive to contribute to creating data spaces through activities such as the Responsible Mobility programme and research into the ethical and societal aspects of such data spaces.

6. Why are you participating in the Data Sharing Coalition?

We believe data sharing is an important topic, albeit still relatively undervalued. The privacy issue often dominates debates about data sharing. But you can also turn it around: we want to be in solidarity as a society (we have the health insurance law for the people who are sickest and we a social system for people who earn less). You can extend this to data, hence data solidarity. I think this concept is a great entry point to engage in a debate. It has to be more than just a privacy or commercial consideration to not share data. Social value is also important and makes it belong to all of us. This stretches way beyond the technological domain of the activities of EAISI, and hence a broader discussion between public organisations, companies and knowledge institutes is required. By joining the Data Sharing Coalition, we aim to contribute to the awareness of this topic, for example, through publication of position papers.

* MPC is also used in two data sharing use cases started by the Data Sharing Coalition. Read more about Improved monitoring of human trafficking and Proving the value potential of MPC for MaaS.

Share:

Read more

SCSN and Catena-X: How to facilitate interoperability between data spaces

SCSN and Catena-X: How to facilitate interoperability between data spaces

In 2023, under the guidance of the CoE-DSC, two operational data spaces, SCSN and Catena-X, explored how end users of either data space could more easily use both data spaces without complex onboarding. What steps are needed to make this possible? What challenges are faced and what possible solutions emerge when going through these steps?