Ordnance Survey releases first fully automated product

30 October 2018

Ordnance Survey has released its first fully automated product derived from large scale source data, which means with the single press of a button OS can create a premium national dataset using the most up-to-date OS source data in just eight days.

In April 2017 a cross-business project team from Ordnance Survey (OS) started work to fully automate the map production of OS VectorMap Local. On 18 October 2018 the new automated derived version of OS VectorMap Local, which is a world leading first by OS, was released to customers.

The OS project team defined several objectives which included delivering efficiencies to OS by replacing the highly manual process of maintaining the product, improving the product currency, and implementing a ‘framework’ system which could be re-used for other OS products. The development also introduced four core principles which were fundamental to the project. These consisted of: trusting the data source (OS’s geospatial database of half a billion features), only using commercial off-the-shelf tools (COTS), creating full end-to-end automation and delivering data quality which meets the needs of the end user.

Jon Butts, OS Project Manager, said: “This is the first fully automated product derived from large scale source data. It has taken a huge amount of effort and persistence from around the business to achieve this long-held aspiration of OS, which is shared by many National Mapping Agencies around the world. “It’s a huge achievement to have developed and created a fully automated derived product, which has a great deal of benefits for our customers and OS by improving the value of the product and increasing production efficiency. “Our customers will see hugely improved currency, consistent refreshes across GB, more analytical value where features no longer end at tile edges, better building parameters, connected roads, and consistency with other products that allows for a consistent zoom stack.”

The team delivered a fully automated map production process solution utilising range of the latest technologies, including Esri ModelBuilder and Python for geo-processing, Esri ArcGIS for publication of map images, as well as end-to-end system orchestration delivered by Workflow Manager. The solution also relies on the Azure cloud environment for data processing.

In addition to the customer benefits the project has resulted in a number of product and operational benefits for OS.

Product benefits - improved product quality

  • Increased currency of data: reduced production time and increased speed to market providing even more value to OS customers.
  • Improved production and product quality through fully automated data processing at National Scale, including : partitioning,  edge-matching, publication and creating change-only updates.
  • Increased consistency of output data  of both contextual and analytical products.

Business benefits – efficiency savings

  • Reduce hardware/support costs through use of COTS software and distributed processing in the cloud.
  • Removal of manual input, freeing up production resource to work on other production flowlines.
  • Use of COTS software removes the cost of developing and maintaining an in-house bespoke capability.

OS VectorMap Local will be released every three months allowing customers to access and benefit from a consistently updated and trusted national geospatial product. As well as producing raster and vector versions of the data, this is also the first OS product to be released in GeoPackage, responding to customer feedback around OS data formats.

The new capabilities delivered by the project were very well received when showcased to the 17 National Mapping Agencies participating in the International Esri Working Group for Multi Resolution Geospatial Production, held in October at Ordnance Survey’s Headquarters, Southampton. Delegates enjoyed two-hours exploring the challenges and outcomes of the project with representatives from the OS development teams.

Website Ordnance Survey

Comments are closed.