Data Governance - Standardise, Process and Deliver
To ensure high-quality, standardised data were delivered to clients, SCIRT's Geographic Information System (GIS) Team created a dynamic process governance model.
In addition to its extensive physical repair works, SCIRT's deliverables included a complete geospatial dataset, showing both newly installed assets and the existing spatial network with decommissioned assets.
Throughout the five-year, $2.2 billion programme, SCIRT's Geographic Information System (GIS) team processed and delivered more than 200,000 spatial assets with associated attribute information to meet client requirements.
To achieve this, a change in mind set over data collection and supply was needed. Each of SCIRT's five Delivery Teams had a different approach. They used individual formats with varying quality. Standards were needed. Not only did SCIRT need data, it needed to ensure it was correct.
To deliver this, SCIRT's GIS team developed and applied a dynamic process governance model, built around a database named the "G-File". The G-File standardised, processed and delivered huge volumes of data that updated back into the client's database for immediate reuse. The G-File database held all the data schema information used by SCIRT for the three waters networks (geometries, feature classes, attributes and values required).
A collaborative process was used from the beginning, taking user requirements and technical specifications into account. As a result, extensive knowledge was built about mapping, particularly between a survey reality and the network models.
A guideline and a template were created from the database, setting out how the five Delivery Teams were to capture three waters' network information. This was extended to supply a guideline and tools for data validation.
The G-File database allowed the SCIRT GIS team to define how best to approach what needed to be captured, and map this information back to how the client held and stored the data. The existing client data structure was not flexible or consistent enough to meet the requirements requested. The GIS team could not change the client's schema, so instead standardised the schema, applying consistent attribute names and enhanced attribute values where required.
The G-File database enabled and controlled automated processes. These processes were developed to facilitate spatial data collection and supply, updates and conversion, quality assurance, publishing and, eventually, handover to the clients. Any changes to schema information were made to the single database and flowed through all processes dynamically.
In summary, the G-File database was an efficient, one-stop maintenance model, capable of providing a high quality and quantity of survey data and managing a complex GIS system to successfully support the SCIRT programme. It was capable of standardising, collecting, processing and delivering.
The G-File model was recognised when SCIRT's GIS team was chosen as a New Zealand Spatial Excellence Awards (NZSEA) 2016 finalist in the technical excellence category. The award application and presentation are attached below.
The key learnings for other organisations are:
- Centralise and standardise: SCIRT's G-File model promoted the centralisation and standardisation of data specifications.
- Quantity and quality: The application of this model at SCIRT significantly improved the quantity and quality of survey data delivered to clients.
- The potential for this model to become a benchmark was illustrated by the fact SCIRT's clients chose to adopt the concept of the G-File model, and it was subsequently used in the definition of new national infrastructure data standards.
- SCIRT's G-File model put data and quality data centre stage, enabling the organisation as a whole to make better, more informed decisions.