Setting the stage: Background and Pilot Case setup
At the end of 2021, a series of workshops convened diverse stakeholders across Flanders, including older adults, local civil servants from five pilot cities, and governmental representatives. The objective of these workshops was a comprehensive exploration of the needs and requirements to enhance the comfort of older adults within urban environments. The primary objective was to comprehensively explore the needs of older adults in urban environments, not only to provide direct assistance but also to equip local policymakers with effective tools for addressing the broader well-being of the ageing demographic.
A significant outcome of these collaborative sessions was the unanimous endorsement of developing an administrative tool to geographically map a particularly vulnerable segment of society: older adults with reduced mobility.
The execution of this initiative hinged on the compilation of two crucial datasets; the age distribution of citizens and the distribution of citizens with reduced mobility. The combination of these datasets aimed to reveal specific zones within the city where older individuals with reduced mobility are concentrated. The analytical process involved the examination of current and historical data, complemented by simulations to predict the potential evolution of this distribution in the future.
This methodology empowers local policymakers and urban planners to integrate the mapped data with existing city service maps, encompassing essential facilities such as pharmacies and other care services. By overlaying these maps, stakeholders can precisely pinpoint areas where services for this vulnerable group are deficient. Armed with this information, policymakers are strategically positioned to plan and develop initiatives that directly address the current and evolving needs of older adults with reduced mobility. Furthermore, they can disseminate these insights to other stakeholders in the ecosystem, fostering collaborative efforts to address persistent needs in a more informed and targeted manner.
When vision meets reality
The collaborative efforts of the Interuniversity Microelectronics Centre (IMEC) and the Digital Flanders Department of the Flemish Government initiated a phased approach, translating strategic objectives into actionable technical tasks. While datasets capturing the age distribution of citizens proved readily accessible from local providers, acquiring data pertaining to the distribution of citizens with reduced mobility posed notable challenges. Several factors contributed to this complexity:
GDPR and Restrictions: Stringent data protection regulations posed a significant hurdle.
Potential Costs: Concerns were raised about the potential financial implications for onboarding, storing, and maintaining data, particularly by the cities involved.
Anonymisation Requirements: There was a potential need for additional anonymisation measures, requiring collaboration with the responsible department of the Flemish Government.
MAGDA Administration: The dataset management under the MAGDA (maximal data sharing between administrations) administration of the Flemish Government strictly necessitates a one-on-one agreement with a pilot city and the Flemish Government.
Despite generating initial interest from civil servants during presentations at various workshops, the aforementioned constraints appeared to temper enthusiasm. Regrettably, no pilot city expressed readiness to undertake the essential steps for onboarding the dataset on citizens' mobility owned by the Flemish Government. In response to these challenges, an alternative, creative, and innovative approach was devised to surmount the impediments and move the project forward.
Pilot case remastered
As the URBANAGE project advanced, prompting the creation of innovative solutions, a fresh perspective emerged for the Pilot Case, leading to the exploration of alternative approaches.
The solution accelerator created for another Flemish pilot case (GCM, Green Comfort Model) is designed to flexibly calculate the Green Comfort Index (GCI) using multiple variables. Combined with the recent development of a configurable index calculator, a new idea popped-up.
We can score the availability of city services for older citizens by mapping the different city service data layers with the citizen’s demographics, creating new indices. This process aims to identify potential hotspots with low scores on the hexagon grid, indicating areas where older citizens may be isolated from essential city services. The calculation method and weights of included parameters can be fine-tuned by local civil servants and city planners to achieve optimal results.
The reworked Pilot case was validated internally during the bi-weekly sync meetings between Digital Flanders and IMEC. This iterative process involved discussions on developments and obstacles, showcasing new features, and conducting evaluations and tests. The initial version of the updated Pilot Case 2 solution underwent thorough internal testing at each step in the development progress.
From “Green Comfort” to “Older Adult Care Service Need” index
We worked out a solution for the city of Gent as a demonstrator. The utilised datasets for this revised pilot case include:
Demographics Data: This data, sourced from the "rijksregister" (national register), encompasses information on the Belgian population. The "rijksregister" serves as an information processing system ensuring the registration, storage, and communication of identification data for individuals. Specifically, the project focuses on the base-distribution population counts for age groups 50-64, 65-79, and 80+ in various statistical sectors within Gent.
Geospatial Index: To incorporate demographics into a geospatial index, the geometries of the statistical sectors in Gent are essential.
Care Facilities Data: recognised by the Flemish Agency (Agentschap Zorg en Gezondheid), which is made available as a POI service on the website of the Flemish Government. The data is updated monthly and originates from the CoBRHA database (Common Base Registry for Healthcare Actors) built by the federal eHealth platform. The Flemish government's agencies of the Welfare, Public Health, and Family domain (WVG) maintain basic identification data of the care facilities they recognise in this database. The following types of care facilities are included:
General hospitals
Psychiatric hospitals
Psychiatric nursing homes
Collaborative sheltered housing initiatives
Centres for mental healthcare
Elderly care facilities
Home care facilities.
Healthcare Service Locations: Locating (commercial) amenities offering healthcare services, such as doctors (M.D.) and pharmacies, is challenging. In Gent, Open Street Map (OSM) data provides a reasonably up-to-date database. The Overpass Turbo function is used to extract data.
Data processing and custom model runs
Following a preparatory round of data processing to get the data formats aligned with the GCM solution accelerator requirements, we merged the resulting open demographic datasets with care facility and healthcare services locations.
Various experimental GCM solution accelerator model runs were conducted. These runs culminated in the creation of the hexagon grid, where each cell is assigned a unique index. This index serves as an indicator, pinpointing the specific requirement for specialised care services tailored to older adults.
The involved datasets are used as individually customisable indicators by adding them to a GCM configuration. For the Gent pilot case visualised below, we use age classes as a weighted indicator for each hexagon; extra relative weight is allocated to the more vulnerable older age groups. Additionally, we integrate the vicinity of care homes to reinforce the index even further.
By employing a colour-coded system on the hexagons, the geographic areas requiring "older adult care service needs" are distinctly highlighted. This information equips city planners and local policymakers with valuable insights.
For other pilot cities in Flanders, supplementary local datasets, like the distribution of individuals with reduced mobility, can be integrated as indicators. This flexibility extends to customising the weights of all these indicators to suit specific needs.
One step beyond - Cascading model runs
The outcome of a Green Comfort model run typically generates a GeoJSON file featuring hexagons with attached scores. These outputs can serve as indicator inputs for subsequent model runs, allowing cities to create composite "sub-indices" that combine and simplify various datasets.
For instance, consider the "older adult care service need" index derived from the Green Comfort model run. Utilising this index alongside other data, such as the locations of general practitioner doctors, pharmacies, hearing aid shops, etc., can produce a new index. This new index specifically targets areas with a mismatch between high service needs and low service accessibility. Consequently, it generates a high gap score, highlighting locations requiring immediate attention to bridge this disparity in service provision.
Furthermore, there is the potential to enrich and reinforce existing parameters like sitting comfort. By aggregating data related to bench locations, bench designs, citizen feedback on benches, and the presence of shade, a new "sitting comfort" index can be formulated. When integrated into subsequent model runs, these comfort scores can contribute to evaluating factors like age-friendliness or overall Green Comfort.
Comments