Data-driven processes require good data management

The water utilities are on the path toward data-driven processes; however it sometimes remains the case that data are not centrally stored, but can only be found spread over various applications, in ‘silos’. It also happens that not everybody within a single organisation works with the same definitions or terms. The fourth 2019 meeting of the Hydroinformatics Platform focused on these questions. The water utilities Brabant Water, Waterbedrijf Groningen and Waternet shared their experiences.

Manage and organise along set guidelines

Hans van Drunen (Brabant Water) explained how Brabant Water uses the international DAMA/DM-BOK (Data Management Body of Knowledge) guidelines to shape its data management and to safeguard data quality. The water utility does this by managing and monitoring all activities that involve data processing in the organisation. The DAMA/DM-BOK guidelines cover ten components:

1. data architectureDetermining and incorporating the data objects and data structures within an organisation in data models.
2. data securitySecuring data against breaches and unauthorised transport inside and outside the organisation.
3. data integrationIntegrating data between applications and reusing data.
4. data qualityFocuses on insight and advice to improve data quality.
5. Data modelling and design
6. data storage and managementDesign, implementation and support of stored data.
7. data warehousingDefining the architecture for data storage in a warehouse, to analyse data and make analysed data available.
8. business intelligenceEnsures the unlocking of data in the data warehouses, to provide information to the organisation’s management for decision-making.
9. meta dataData about data, in the form of organisational terminology and definitions.
10. reference and master dataMaking data easily findable through standards, definitions, integration, management, and by determining and incorporating authentic sources.

It is important that all these components of data management be designed and carried out in coordination. Implementing data management systems is challenging. For instance, attention has to be paid to how users make the system their own. Moreover, clear agreements are required when it comes to responsibilities, for example, about who is the owner of what data. Water utilities must also invest in raising their employees’ awareness that data constitute an important asset. But nobody is alone is this effort. Many standards have already been established, and many other parties, like municipalities, Rijkswaterstaat, Rioned and the Cadastre, have already made advances and can serve as examples.

Data warehouse delivers to everybody, from operator to manager

Gijsbert Haaksman (Waterbedrijf Groningen) then made a presentation on how Waterbedrijf Groningen’s data warehouse ensures that data are available to everybody, from operator to manager. Waterbedrijf Groningen, WMD (drinking water supplier in the Province of Drenthe) and WLN share infrastructure, data sources, software packages, as well as knowledge and expertise. This makes data structuring a key condition for efficient collaboration. This is why a data warehouse was set up, using the Tableau software package as a visualisation tool. The Groningen water utility’s whole system has been built and the reports in the system are automatically updated. But even if the system is already operational, it remains a ‘work in progress’. The advice of those involved: data warehousing projects are big, so begin small.

Linked data

Paul Stroet then showed how linked data are used within the digital transformation at Waternet. Linked data is a digital method for the publication of structured data, in a manner that these can be made available on the internet and thereby also become more usable. This is achieved through the use of unambiguous terms and definitions, the establishment of mutual interactions and invariable web addresses (URI: uniform resource identifier).

Data-driven processes require some ‘Data Literacy’ from the employees. This is important, because sometimes several different kinds of jargon are used within a company. For example, there may be various terms for a single object like a pumping station, or terms may not be clearly defined. A structure needs to be created to effectively share the data internally, but also to make the organisation a better chain partner in communications with other entities, such as municipalities, the Cadastre, Rijkswaterstaat, and with contractors or other companies.

Lessons learned and advice from the frontrunners

This year’s fourth meeting produced the following recommendations:

• You have to start somewhere, so, just do it.
• Recognise the importance of trainings, and take the time to involve different levels and sections within the organisation in data management.
• You must give data management the time to grow; make the appropriate time and personnel resources available.
• Look for a sponsor on the MT and make links to existing projects; find the right leader.
• Be aware that this is a learning process. It is valuable, but it can be laborious.
• Talk about data management can become very abstract, therefore focus on results, not on the concept.
• Link up with the ‘pioneers’ in data management.

Hans van Drunen of Brabant Water shares his company’s experiences.