National Mapping Agencies such as Ordnance Survey Ireland need to strategically reposition themselves over the next 10 years by taking advantage of technological advances to drive new ways of collecting, curating, and delivering geospatial data. OSi first encountered the ADAPT Centre and its research through a professional networking event. The first engagement took place in 2016, with further projects taking place every year since, each new project building upon the last while expanding the scope of the collaboration. The initial goal of OSi’s engagement with ADAPT was to develop a platform to publish their geospatial data as Linked Data on the Web while adhering to best practices in the domain of geospatial information but has broadened into addressing challenges in areas of data quality and governance, as well as in image processing.
ADAPT’s Linked Data experts in conjunction with OSi’s geospatial experts developed OSi’s Linked Data Platform (http://data.geohive.ie) that makes available Ireland’s geospatial data according to Linked Data principles. The adoption of Linked Data enables third parties to explore and consume rich data in a meaningful manner via a combination of simple, standardized technologies (e.g., RDF and URI) that operate over the Web’s existing HTTP infrastructure facilitating intelligent integration and interaction with data. The knowledge engineering of the RDF model, the uplift of OSi’s Prime2 relational database data into RDF, the design of appropriate URI strategies, and the development and deployment of the Linked Data Platform (http://data.geohive.ie), were some of the tasks in which ADAPT’s expertise was used in supporting OSi to meet their open data publication challenge. Data.geohive.ie allows third parties to consult authoritative administrative boundary data, selected data about buildings and enable the development of novel applications on top of these.
To support the adoption of the approach, a demonstrator app has been developed as an example to App developers showing how they can interact with and benefit from OSi’s linked data.
Also ADAPT and OSi have ollaborated on ways in which image processing and machine learning can be improved to identify map features (such as buildings) from aerial imagery and use other sources of information (e.g. raw DTM, shadow, etc.) to accurately geolocate the features and extract information from raw data that can be used to position objects in 3D. This image analysis using statistical and AI methods to determine geospatial features could play a role in avoiding frequent physical resurveying.
Research has also been performed to develop a prototype solution for access control for “Closed” Linked Data, and a prototype tool for the conversion of OSi building geospatial footprints into Linked Building Data, with a focus on the creation of rudimentary BIM models which include wall elements.
As a result of these projects with the ADAPT Centre, OSi expects that web developers will now be able to interact with some of OSi’s key data sets and content in more intelligent, exciting and creative ways, allowing for more meaningful interactions with the data and allowing OSi to create new commercial products and services.
In response to the COVID 19 Crisis, OSi and ADAPT enabled access to COVID-19 Health Surveillance Monitor (Ireland) data as open data in the high quality (5-star) Linked Open Data format through the data.geohive.ie portal at http://data.geohive.ie/covidData.html and through data.gov.ie.
Continuing on from these successful collaborations, and on research already outlined in the OSi strategic plan for 2019 to end 2021, some further challenges are currently being looked at in this next stage of engagement.
The first of these challenges is in the establishment of an appropriate data governance framework, cognisant of existing practices whilst endeavoring to integrate state of the art approaches and technologies to improve the governance of the entire data lifecycle (driven by the key quality of service imperative-authoritativeness), from generating the geospatial data to begin with (e.g. through sensing, image processing etc.) through quality assessment and control through delivery and ongoing maintenance.
ADAPT researchers discuss Ethics & Emotion in AI at @RCSI_Irl and @TrinityResInnov Innovation Showcase 2021: https://bit.ly/3I4czEP #ResponsibleAI