This exclusive interview was published in issue 106 of Position magazine, published while Australian authorities were initially scrambling to contain the spread of Covid-19. Gary Costa of MapData Services had been doing some scrambling of his own – to develop a unique, robust dataset to inform the effective response to the pandemic.
DB: Gary, we appreciate your time to participate in these extraordinary circumstances. I understand that MapData Services has been working around the clock to develop datasets to inform response to the Covid-19 crisis. Could you briefly describe the core needs you were responding to, and the process of developing these resources to date?
GC: These truly have been extraordinary circumstances – and really highlighted for many just how critical having access to the right data is when it comes to making effective and informed decisions. What we’ve seen as the Covid-19 situation has unfolded, is that the challenge for the community and government alike is the shortfall of authoritative data. In the few months since this crisis began, there has been no single dataset available that provides a current and comprehensive perspective on the Coronavirus situation in Australia. Each state has their own data of course, but this is all stored in disparate files and systems with core differences in the type of information available between jurisdictions.
We saw this gap and understood how critical it was for decision-makers in any industry — whether its government or commercial entities — to get their hands on a national Covid-19 dataset. It’s not uncommon that there is a need for a dataset that doesn’t exist yet. Identifying these data needs and filling these gaps is a key focus for the team at MapData Services.
To create the first comprehensive dataset of the known Covid-19 locations in Australia we partnered with Early Warning Network (EWN) to source, verify and map the information into a format that would be consumable – and valuable – for government entities, commercial organisations and the community at large.
Outside the COVID-19 Local Cases dataset, we also have a broad range of other data that has proved crucial to understanding the associated implications of the virus. For example, as social distancing has become a reality, it’s been important for officials to understand human movement and how this may have changed as new measures have been put in place.
Our human movement data is collected from mobile phone apps. It anonymously identifies the device from which the location record originated, consistent over time for a given device, which allows us to study aggregated movement of those devices. Insights you can glean include stay/linger time, local population movements, and travel distance.
We also have traffic data, which, when mapped and analysed against all these other layers of information, can also provide compelling insights into how people are responding to the virus.
DB: What is the infrastructure supporting these resources, and what are some of the types of analyses this facilitates?
GC: The COVID-19 Local Cases dataset has been compiled based on media reports and government updates, to provide a national view of where cases are occurring with specific details about those cases. This is an unfolding situation, with discrepant data typically held between states and territories, so the aim here is to provide a national view that is as complete and up to date as possible.
EWN has a team of data specialists who are checking and verifying the information before adding it to the map. If it’s just an individual report, a small polygon is placed at the location where the case has been reported. If there are a number of cases in one location, we draw a larger polygon.
From an infrastructure perspective, we’re using ArcGIS and FME software to integrate all the data and translate it into a consumable format including as an Esri Feature Service, GDB, ShapeFile or GeoJson file.
DB: Who are the key stakeholders in terms of end users?
GC: There’s been a real demand for this kind of data across a broad range of government sectors, particularly those with a national outlook. Beyond that, banks have taken a keen interest in the COVID-19 Local Cases data, with one of the big 4 banks recently undertaking some exploratory work using the dataset to understand which of their branches are most impacted by the virus. Health services, such as doctors’ surgeries, have also looked to the data to understand what’s happening in their local area.
DB: What are the most critical use cases?
GC: There are many ways this data can be used to support informed decision-making. I think understanding hot spots to enable decision-making is the most critical use case. You can clearly see where outbreaks are occurring across the country, and also access details on reported cases, including date, location and the number of people affected. You can discover how the spread of the virus may impact delivery services and assign appropriate medical supplies and health services to areas in need. You can also make data-driven business decisions on service continuity – something many people have been dealing with on a large-scale.
DB: How do you anticipate these changing as the virus spreads and the numbers of casualties increase?
GC: Information is power, and like in any situation, having access to more data means better decisions can be made.
DB: How might these resources be expanded in terms of partners and capabilities in future?
GC: We’ll continue to work closely with our partners to constantly refine what we’re doing and expand the types of datasets available. As I said earlier, it’s not uncommon for organisations to want access to data that doesn’t yet exist – and our job is to work with our global network of content creators and data custodians to source, build and create those datasets and make that information available.
DB: Will this data be made available to the public via web interface or an API?
GC: For organisations looking to add this data capability into their systems, it can be accessed via an Esri feature service or a feed into your system in your required GPS format. Our team can also assist you to integrate it into your existing systems. We’re also working to have some freemium versions of this data shared publicly via the Australian Covid-19 Hub, available at: www.covid19-esriau.hub.arcgis.com. This Hub is a really great resource for anyone wanting a starting point for data relating to Covid-19 in Australia.
DB: Are there any ways that our audience of spatial professionals may get involved to assist with developing resources in the ongoing response to the Covid-19 outbreak?
GC: I encourage anyone who has data that may assist during this crisis to share it via the Australian Covid-19 Hub. This resource is being widely used by many government agencies and already features free data from Guardian Australia, ABS and others.
As you’ll see in the Hub, there’s already lots of valuable, complimentary content you can access online – from government open data portals or public websites – however to undertake really valuable analysis, you typically need to complement this with specialised data. If there is data that is needed but doesn’t yet exist, please reach out to the team at MapData Services. We have an extraordinary network of data partners – from across the country and around the world – and can mobilise to source any kind of data required. Having access to the right data really is critical – especially when dealing with a pandemic of this scale.
DB: Many thanks for your time Gary, and good luck.
Stay up to date by getting stories like this delivered to your mailbox.
Sign up to receive our free weekly Spatial Source newsletter.