From here to there: How the British Red Cross uses proximity to support our services
Have you ever wondered how organisations like the British Red Cross determine the accessibility of their services or the suitability of their locations? The intention of this blog is to show how analysing location and geography can contribute to answering questions like these.
As the data literacy of the organisation grows, users are becoming more accustomed to using geodata and geo tools in their day-to-day work. Their increased appetite and expectations mean location-related questions are getting more complex.
One of the ways we show people the value of spatial analysis is through our weekly interactive drop-in sessions, which are rooted in a community-focused surgery approach
One topic that often pops up in these sessions is ‘proximity’ — but that single word belies the potential complexity of possible questions and scenarios about how close something is to something else. Some look at it in terms of travel time between destinations, others by geographic areas they can access from a known location, and some a straight-line distance. Each question has its own operational context and audience and therefore needs specific tools and methods to be applied to it.
To make things clearer, we’ve gathered some user stories in this blog, to showcase the different aspects of proximity analysis. The blog is both an overview for our community of the main types of analysis and a way to share with others what we’re working on in the Red Cross geo-world.
Understanding Distance, Time, and Reach
To start with, here are the three most common things our users want to know and a map showing possible answers to these questions:
- How to visualise the area a certain distance away from a location
- The time required to move between two set locations
- The furthest area you can reach from a location in a given time or distance
Understanding User Stories
Before we dive into examples of proximity analysis through three user stories, let’s take a moment to understand what a user story is. At its core, a user story is a concise, plain-language description of one feature of the mapping product, described from the perspective of an end-user. It follows a simple formula:
“As a [type of user], I want [an action] so that [a benefit/a value]”.
These user stories help to convey how users interact with a map product and what they aim to achieve, ensuring that the focus remains on user needs and experiences. For more information on user stories have a look at this page from Atlassian and this document from ESRI on the same concept for developing geospatial products.
User Story 1: Mobility Aids Service
“As a Mobility Aids Service (MAS) hub manager, I want to visualise a 20-mile straight-line buffer around my MAS Hubs on a map, so that I can easily determine which other public services are within that 20-mile radius.”
- To complete this analysis the locations were added to the map and the buffer tool was run, creating geometric circles on the map.
- These shapes were then used to filter other locations that were either inside or outside the buffer.
- The map below shows 20-mile buffers created around some MAS Hubs in north Wales and Wirral.
- In this case, the data is shown on an interactive map in Microsoft Excel.
User Story 2: Emergency Response
“As an Emergency Response Coordinator, I want to know how long it would take for an ER volunteer to travel to the home base of an ER vehicle, so that the ER team can respond to an incident in the most efficient way.”
- This analysis created a table with pairs of origin and destination locations and the journey times between them.
- To complete this, the location pairs were added to the map, then the analysis tool run with the other variables required.
- (Note: by default the tool calculates the distances/times between all origins and all destinations, but for this use case single defined pairs of origins and destinations were used. Although this sounds simpler, this requirement made the analysis process significantly more difficult to implement).
- The table output was further analysed to show that, for example, 50% of ER volunteers live within 30 mins of an ER vehicle.
- Below is an example of a table created with this analysis (Note: this does not show real ER data and nor are the figures quoted above real).
User Story 3: Refugee Support
“As a Refugee Support service manager concerned with working with asylum seekers housed in temporary accommodation set up by the government, I want to visualise the services that are within a 20-minute walk from the temporary accommodation, so that I can understand and assess the potential needs of someone housed there.”
- This analysis created a set of isochrones — these are polygons that show the geographic area accessible from a point on a network, within a certain time or distance threshold (isochrones are also known as “travel time areas”, “service areas”, “catchment areas” and “drivetime areas”).
- The starting locations were added to the map, along with the network dataset. The analysis tool was run on those locations and the isochrones created.
- These were then used to select and filter other locations on the map to represent what is in range and out of range of the starting locations, to provide a measurement of accessibility to services for each location. For example, the service manager was able to compare the accessibility of large supermarkets between the Temporary Accommodation Sites analysed, so that they could determine whether or not supermarket vouchers were a suitable support mechanism for people housed at the sites.
- The outcome and impact of this analysis is described in much greater detail by my esteemed IM colleague Alex Pycroft in his Medium blog: “How we used data to improve dignity and choice in our services”
- In the map below, the coloured shapes represent locations of relevant buildings, the points represent the locations of bus stops and the coloured lines show estimated walking times from a starting location (Note: this does not show real Refugee Support data, although the Ordnance Survey and Department for Transport data is accurate)
Adding complexity
Depending on the use case and the tool being used, it may be technically possible to add additional variables or constraints into the analysis to, for example, improve accuracy or repeatability. However, there is a balance here: is the value of the new answer given worth the extra complexity? (i.e., what is “good enough”?)
In the analysis itself, variables to consider adding might be:
- Estimating the change in time or distance based on the start time and/or day of the journey. For example, a journey travelling through Birmingham starting at 0800 on a Monday morning, may take significantly longer than the same journey starting at 0100 on a Sunday morning. This may make a difference to something like an incident response time for an ER team.
- Varying the travel mode to more accurately reflect the type of journey the average user would have or may be prepared to undertake. Depending on the tool, travel modes can model the duration of journeys for small vehicles, large trucks, cycling, public transport etc. For example, a travel duration (and therefore the effort required) for someone walking a mile would be different to someone driving or taking public transport. This would be important if you were analysing how far someone would be prepared to travel to access a key service, like a source of safe drinking water (See: Sphere Standard 2.1 Access and Water Quantity)
- Journey times for different travel modes can also be affected by constraints within the network. These are sometimes built into the network itself (e.g., a speed limit, or a low bridge), or can be added temporarily (e.g., a known area of flooding which is blocking a road). Again, depending on the use case, these types of constraints can affect journey durations which may, for example, have an impact on things like logistics for vaccine delivery
The use case helps guide the complexity required to run the analysis too. Some questions that may be considered in the design stage are outlined below (note: many of these questions apply to other types of analysis too).
What should the end product be?
- Where and how is the data stored and where and how will the end user access the result of the analysis?
- Is the data a table of locations to be analysed all at once or will a user enter a single location themselves and receive one result?
- Is the product for one user or many?
- Should the product be a static map or interactive app? A data export to analyse in Excel or combined with other data for display in PowerBI?
- Is the data and analysis sensitive and internal or public-facing and advocating?
- Is the analysis product tactical or strategic?
- If the initial product proves useful, how could it scale to fulfil the needs of other users?
Is this analysis a one off or does it need to reflect data that changes every day/week/month/quarter?
- E.g., if an analysis is run to show the journey time for every staff member to their designated office on the HR system, is a one-off analysis sufficient? Or should the process be run periodically via a script on data integrated from other corporate systems, perhaps in a secure virtual environment.
What’s the budget?
- For the analysis it’s possible to use online services which are quick and simple but cost money to use (e.g., ArcGIS Online) or use free datasets and run the analysis locally (e.g., Open Street Map or Ordnance Survey Open Roads) — but this takes longer and needs the dataset to be prepared correctly.
- It’s also possible to use alternative online sources for analysis which sometimes have free tiers for non-profits (e.g., Traveltime), which integrate into QGIS and ArcGIS Pro
How many locations are being analysed and where are they?
- Analysing the proximity of a single staff member to their office in the UK, is significantly less complex than, for example, analysing the accessibility of health clinics in Ghana. The latter analysis is both more computationally intensive and a comprehensive and topologically correct network dataset for the area potentially harder to access or create.
Does overlap in journeys matter?
- If a destination is reachable from several different origins, should the amount of overlap be analysed more? Does a lack of overlap indicate something notable? For example, if analysing the theoretical incident response times of lots of Emergency Response teams, does having lots of overlapping coverage in certain areas, and just a single team (or none) able to reach other locations matter? Is the organisation able to reach those it needs to reach? Is having overlap a good thing for redundancy?
- In the map below, multiple journey time areas have been stacked on top of each other and then the total number has been aggregated to a hexagonal grid. This gives a generalised overview of the “reachability” of different areas from the origin locations. For more information on this take a look at this article by Kontur and the DisasterNinja application
Conclusion
Proximity analysis is a powerful tool that can aid us to understand where things are in relation to one another. In the humanitarian sector, as the complexity of location-based inquiries increases, it becomes ever more crucial to understand the ‘why’ behind each question, so as to ensure appropriate subsequent analysis. Hopefully this overview of proximity-related user stories provides a clearer understanding of the considerations involved in this type of analysis.
It is worth highlighting here that geospatial analysis can only ever give a geospatial answer, and whilst geography is an important component of many questions, it is only one factor.
It is also worth highlighting that for public facing, advocacy-type products, where user experience in terms of accessibility, platform performance or UX/UI is priority, then we would defer to the expertise of our colleagues in Digital and Communications — it may be that we can provide a geospatial contribution, but those teams would be best placed to guide on a end-product.
If you are reading this in the British Red Cross and thinking “I too have a question about location and want to analyse my data geographically”, then please drop into our weekly surgery space and we can have an informal chat — the details are on our RedRoom page. Alternatively, if you’re in the Red Cross movement feel free to reach out to your National Societies’ GIS or Information Management colleagues and look out for the Surge Information Management Support (SIMS) community of practice which includes a number of our GIS colleagues and reaches across borders and time zones.
