Björn Ottersten, director of the Interdisciplinary Centre for Security, Reliability and Trust at the University of Luxembourg, talks to Funds Europe about fintech research and applying these projects to the real economy.
There is a practical feel to the research work conducted in the University of Luxembourg’s security and trust (SnT) centre. Founding professor Björn Ottersten explains that the centre was created to drive the transfer of research into the Luxembourg economy.
“One of our hallmarks as an academic institution is that we are heavily focused on translating technology into applied applications,” he says. “We aim to attract talent from around the world, to provide training and research and to fuel the economy by developing talent in areas where there is a shortage of this set of skills.”
The focus is on facilitating public-private partnerships: on using public investment to encourage private investment in the financial sector and in other areas of the economy.
A slow burner
Initially, the partnership between the financial sector in Luxembourg and this specialist university research unit took longer to grow than expected. “When I first came to Luxembourg to establish SnT, I had assumed that we would be working closely from the beginning with the financial sector, given the importance of this segment to the Luxembourg economy,” he says. In reality, the initial interest of the financial sector in technology collaboration was lower than it is today – and, for this reason, SnT stepped into partnerships with companies from a range of disciplines, including SES, now the world’s largest satellite operator, Lux Post, Telindus, and prominent data centres and providers of cloud computing services.
More recently, the financial sector has started to take a more dynamic interest in fintech research and the benefits this can deliver to its operating efficiency and profitability. “Roughly five years ago, there was a tangible rise in interest from financial companies and we could really start to talk about developing project partnerships at the seed level,” says Ottersten.
This dialogue has translated into projects in four main categories:
- Distributed ledger technology (DLT), where SnT is running a number of projects with financial sector partners;
- Cyber security, which is becoming a priority for the financial sector and critical to the stability and resilience of the digital financial ecosystem;
- Data analytics, including the application of artificial intelligence (AI) and machine learning to the large body of transactional data that financial organisations hold in Luxembourg; and
- Compliance and regulatory technology, which involves the application of technology to help firms adapt to regulatory change and comply with these obligations.
In the cyber-security area, some firms are stepping beyond perimeter-based security to explore multi-layered approaches to their network and database security. This trend is becoming increasingly data-driven, with firms utilising the large data sets to which they have access, often in combination with artificial intelligence techniques, to identify abnormalities and risk patterns that provide early warnings of compromised security.
Telindus, the IT and digital solutions arm of Proximus Luxembourg, has been working with SnT to develop a security dashboard that will provide early warning of suspicious or abnormal patterns. These techniques may be applied to structured data (for example, data held in a relational database), but also unstructured data such as text drawn from social media channels or news feeds. Improvements in natural language processing (NLP) have been important in enhancing the industry’s ability to identify risk indicators from free text.
Requirements engineering
In applying AI to process automation, the aim is not to target 100% automation immediately. Typically some human intervention may be used to verify the accuracy of the task fulfilment and to apply a learning component, so that the error rate reduces during the training process. By doing so, this allows the algorithm to fulfil the task requirements with greater accuracy and for the human intervention requirement to be gradually reduced.
This AI or reinforcement learning approach is also being applied to developing compliance software. Working in partnership with the Central Legislative Service (SCL) and Digital Luxembourg, a team at SnT is using NLP to extract compliance requirements from legal texts and to transform these into structured rules. Researchers are applying techniques from requirements engineering – the process of defining, documenting and maintaining requirements in engineering design – to drive this. Using formal methods, the work allows reasoning and checking of regulation as well as automatic testing to ensure compliance. Ottersten says that this type of compliance technology is not only being applied within the financial sector: SnT has recently signed a four-year research agreement with a legal firm to develop internal tools that will monitor its compliance with GDPR.
It is also working with insurance companies to develop automotive contracts where the premium is aligned to the driver’s risk profile. The company will monitor the customer’s driving over a period and will utilise behaviour recognition software to evaluate risk and to offer customised pricing.
Blockchain and governance
DLT is being used alongside AI mechanisms such as those described above. SnT researchers are exploring the use of blockchain technology to develop shared and trusted know-your-customer solutions, taking advantage of the fact that information on the blockchain can be amended only with the majority consensus of participants across the blockchain network. SnT has also established a partnership with a Luxembourg company that uses blockchain technology to bring donors and beneficiaries together in the area of Islamic finance.
Applications of DLT have presented challenges around scalability. “By its nature, software is always buggy,” says Ottersten. “When a bug arises in a decentralised blockchain framework, this may result in a division or fork in the community, with one group favouring changes to the code to resolve this issue and others wanting to continue with the existing codebase.”
In Luxembourg, a project called InfraChain has been developed to address this challenge, providing a governance structure around distributed ledger applications and a legal mechanism for dispute resolution.
Beyond the financial world, SnT has been working extensively with SES, the major global satellite services provider. The satellite business is also undergoing digitalisation, radically changing the design of satellites and the operation of communication networks. The use of a large number of low-orbit satellite systems is being evaluated to support communications and data services across the globe. This strategy is driven by measures to improve reusability of satellite launch technology, thus reducing the major financial and resource cost involved.
©2019 funds europe
