Beltug

Information ownership determines the speed of your data transformation. Takeaways from the Beltug N-sight: 21 October 2020


In this era of digitisation, data is often called the 'new gold' or even the ‘new oil’.  As we seek to gain more insights that will lead us to higher revenue, new market opportunities or new regions, we are analysing data at full throttle. But this data needs to be handled with care, using a data architecture that follows your general strategy while ensuring solid security, quality, etc.

 

During this session, we heard from companies about their real-life experiences and best practices: Borealis set the scene with challenges; Brussels Airport Company and AXA Bank shared their vision of data governance; and Stibbe gave us the legal perspective for ‘non-personal’ (i.e. 'industrial') data.

 

Presentations from the event are available to our members (after log-in).

 

 

Case: Borealis' data lake journey

 

Data is a hot theme, and Borealis is no exception, explained Nathalie Rigouts, Head of IT Innovation.  Four values guide the business decisions of this globally present company, and drive every action they take:

 

Borealis has made quite some progress in its data journey, yet there is still work to be done when it comes to predictive and prescriptive data (slides 8 and 9). A Data Platform has been created in the last year as a way to make sure that all data-related profiles can rely on the same set of data, rather than different data silos.

 

People need to be able to access this platform without learning new skills. And the platform needs to avoid redundancy, so everyone can benefit from the work of others. Finally, the success of the data platform is measured by delivering access (the output) rather than the input (collecting as much data as possible).

 

One key element is the data lake, which is grown incrementally, in an iterative way, driven by business needs. In this data lake, Borealis aims to introduce fundamental capabilities for re-use, automation, scaling and efficiency (slides 12 and 13).

 

The Data Warehouse, however, also remains an equally important element, with Borealis combining both worlds (slide 14).

 

Nathalie emphasised several ‘do’s and don’ts’ for data at Borealis:

 

To wrap up, we learned from Borealis' experiences:

 

 

It's not personal: the legal perspective on industrial data

 

Next, Erik Valgaeren, Head of the TMT & Data Governance practice at Stibbe, threw a light on the legal perspective of data. To start with, Erik wondered: what exactly is data? Is this a question of bits, or records? Are we only talking about structured data, or unstructured data as well? What about ownership of data: people often assume that they own all data that resides on their systems, but do they?

 

The truth is that different legal regimes may apply simultaneously to a certain set of data. Often, it's the contracts that stipulate what can and can't be done with the data. For organisations such as banks, hospitals, etc., a lot of the data being handled is subject to sector-specific legislation. And certainly, there is the touching point with GDPR when it comes to personal data.

 

Be aware of defining those personal data in a too-narrow way: it covers much more than home address, email address, name, telephone number, date of birth, etc. Furthermore, at the end of 2018, the EU issued a new regulation - this time on non-personal data.

 

When it comes to data governance, make sure to use clauses in your contracts with suppliers and business partners, etc., with

 

Securing your data, regular back-ups and solid access controls are also part of a proper data governance.

 

To wrap up, Erik shared a use case about a provider wanting to (re-)use your customer’s location data for his own analytics and statistical purposes, and offering remuneration (slide 13).

 

 

Case: How data governance took off using business needs as main driver

 

Joke Op den Acker, Operational Data Manager at Brussels Airport shared with us that 'customer first' and 'passenger experience' are the main focus points for the airport, as it aims to create value for stakeholders.

 

In its own journey, Brussels Airport has moved from reactive to proactive and even to predictive data management. The complexity of the hub always needs to be taken into account, and the facilitating role between internal and external partners is vital.

 

An airport is an ecosystem and the passenger consumes the services of that ecosystem which includes handlers, airlines, air traffic control, commercial partners and 3rd party providers (slide 6). The passenger journey needs to be coordinated end-to-end for: accessibility, passengers, luggage and aircraft.

 

Over the last 10 years, Brussels Airport has worked on an 'Airport Operations Plan' (AOP). In 2010, it began sharing data with partners in the ecosystem to allow for the first steps towards predictive data management. Then, in 2015 they moved to the APOC (Airport Operation Centre) - the heart of the airport, where several decision makers were brought together to decide upon flight related mitigation actions based on shared data, collaborative decision making and a lot of mutual trust and joined forces.

 

And finally, in 2020, the AOP will aim at even more prediction and forecasting models and putting the learning back into practice. In the long term, this AOP will be a daily guide, steering operational and business decisions, based on a 'PLAN-DO-ACT' circle (slide 9).

 

HELI is the 'face' of the AOP. It integrates the planning into one platform and visualises the data that are analysed and interpreted in the AOP. Via HELI, all stakeholders in the airport can easily track, monitor and manage information about the overall airport operations. Brussels Airport is gradually integrating machine learning modules, data science and artificial intelligence to evolve into an application that can suggest actions to its users.

 

Joke shared an example of a forecast capability, with the prediction of flight delay patterns (slide 12). This can help airlines make business decisions: proactively re-booking passengers, starting handlers to work on available aircraft instead of waiting for the delayed one, etc. And there are plenty more examples (slides 13 and 14).

 

Data governance is carried out by the Data Stewards team. For the moment they mainly focus on the data flows for operational flows ('passenger baggage' and 'aircraft and flight'). Based on 3 pillars, Joke can state that Brussels Airport is now a data-driven organisation:

 

 

Case: The Happy Data Triangle

 

Our final real-life case came from Jeroen Ghysel, Chief Retail, IT & Transformation at Axa Bank. For Axa Bank, unlocking the value of data is not merely a slogan. They go all out for data to:

 

Four years ago, Axa Bank began an 'as is' analysis. It included observations such as: what data did the bank have available, where did the data reside, who owned the data, who was responsible/accountable for the data quality, and what about the 'how' dimension: exploiting the data.

 

Jeroen shared the 'Happy Data Triangle', with 'Who' on top, 'Where' and 'How' at each angle of the base, and 'What' in the body of the triangle.

 

'What' covered the creation of clear definitions throughout the company on the activities of the bank. One of the missions of this data journey was to bring data to the entire company: so Jeroen created a Data Board, with the CDO, the MB -1's, the Head of Security & Privacy, the Chief Architect, the Head of Data Management and the Head of Data Governance. This board meets every month to steers the data strategy for Axa Bank. The start was a bit challenging, as the people in this board weren't very familiar with the concept of data and data science.

 

This Data Board is a vital part of Axa’s combined bottom-up/ top-down approach:

 

For the ‘Where’ aspect, the question is 'where is the data I need'. Are data lakes and data warehouses the right approach? Axa chose a use case-based approach. Rather than defining a target data architecture, it works with key principles. Every new project/investment needs to respect those principles.

 

The last part of the triangle, is the 'How'. End-users at Axa Bank have self-service tools at their fingertips now, with standard reporting tools, analytics, a data lab and AI tools. All of this increases the autonomy of the business, which, after all, has a key role in exploiting the data.

 

The key to success, Jeroen emphasised, is to start with the end vision in mind… but dare to take one step (use case) at a time. And build a maturity assessment for all of your data capabilities - then repeat it every year.

 

Axa Bank's key takeaways (so far) are:

 

 



 

Dear visitor,

Access to more information about this topic and/or to download the paper is easy and fast, but exclusively for Beltug members (just login to get access).

Beltug gathers a lot of information. Here you find the advantages of Beltug membership

The Beltug Team

Click here to login