Cloud mapping without the cloud: what one bad Wi-Fi network taught us about capability

The 2017 Indigenous Mapping Wānanga at Claudelands in Hamilton was built around modern mapping tools, practical training and a large collaborative programme. It also provided a very effective demonstration of something much less glamorous: the cloud only works if you can actually reach it. A large room full of participants and trainers opened web mapping applications, imagery and online services at roughly the same time. The venue Wi-Fi struggled, connections slowed and trainers began improvising.

It was frustrating at the time, but it remains one of the better lessons I have taken from years of GIS capability work. The problem was not simply technical. It exposed the difference between a technology being theoretically available and people having a realistic, reliable way to use it for their own kaupapa.

The technology industry often talks about cloud services as though they have escaped infrastructure. Data is “in the cloud”, applications are “cloud native” and users can supposedly work from anywhere. All of that can be true, but the cloud still has to arrive through fibre, routers, wireless access points, devices and whatever security settings stand between the user and the service.

At Claudelands the tools existed perfectly well somewhere else. The problem was that the local environment could not deliver enough connectivity to the people trying to use them. That distinction matters because access is often described too simply. A software product may be free, a dataset may be public and a service may be available nationally, but none of those things guarantees that the person who needs it can actually use it.

This matters in Māori GIS because capability is not only about giving people a login to somebody else's system. It is about whether iwi, hapū and other Māori organisations can choose tools that fit their circumstances and still exercise meaningful control over the work. A cloud-first design may be excellent for one kaupapa and completely wrong for another. A solution that works brilliantly in a central government office with fast fibre may be a poor fit somewhere with weak connectivity or where information is intended to remain local.

The wānanga continued because trainers and participants adapted. People changed exercises, shared connections, slowed the process down and helped one another. That response was as important as the failure itself. The technology was not the only source of capability in the room. Whanaungatanga mattered. People were able to keep the learning moving because they were working together rather than each participant being treated as an isolated user of a platform.

That experience also reinforced the value of ako. Trainers were not simply delivering a fixed programme to passive recipients. The conditions forced everyone to learn from what was happening in front of them. Participants brought their own constraints and workarounds. The programme had to respond to the reality of the room rather than insisting that the original plan was correct because it looked tidy on paper.

There is a wider lesson in that. Māori GIS capability should not be designed around an idealised user with unlimited bandwidth, a new laptop and full administrator rights. It should begin with the people, places and constraints that actually exist. The kaupapa should shape the architecture.

That is one reason offline-capable tools and local data remain important. QGIS can continue to be useful without an internet connection if the required datasets are available locally. Mobile collection systems can be configured to work offline and synchronise later. Basemap packages can be prepared in advance. Critical information can remain under local control rather than depending entirely on the continuing availability of an external service.

This does not mean rejecting cloud GIS. Cloud tools are extraordinarily useful for collaboration, publishing and access. The question is whether the technology increases capability and rangatiratanga or merely creates another dependency. If the organisation cannot export its information, does not control the account or cannot use the system when connectivity disappears, then the apparent convenience deserves another look.

There is also an intergenerational question. A system may be funded for one project, but the information can matter for much longer. Kaitiakitanga in this context reminds me that technology choices create responsibilities to those who inherit the work. Is the information still understandable if the platform changes? Is there a local copy? Does anyone know where the source came from? Can the next group of users continue without rebuilding everything from the beginning?

The Claudelands Wi-Fi failure was funny in the way technology failures become funny after enough time has passed, but the lesson has stayed useful. Access is never abstract. Somewhere the cloud still arrives through physical infrastructure, and somewhere a real person has to make the system work in their own environment.

Good capability building therefore starts with a less glamorous question than “What can this platform do?” It asks, “What do these people need to do, under whose authority, in this place, with the infrastructure they actually have?” Once that is clear, the technology can take its proper place as a tool rather than the centre of the story.

For the wider history of the Indigenous Mapping Wānanga and Māori GIS practitioner networks, see From conferences to wānanga.