For VU.CITY, that matters because our customers do not use spatial data in abstract settings. They use it in planning committees, investment cases, public consultations, infrastructure reviews and design conversations. These are places where evidence is challenged, and rightly so.
A visualisation can be persuasive. But persuasion is not enough. People need to know what they are looking at, where it came from, when it was last updated and whether it can be relied on. That is where provenance matters.
Provenance sounds technical, but in practice it is simple. Every data layer should be able to answer four basic questions:
Without those answers, trust breaks down quickly. Definitions drift. Layers fall out of date. Different teams work from different versions of the same evidence. A map may look clear on screen, while the confidence behind it is much harder to judge. For spatial data, that risk is particularly important. A planning application, infrastructure assessment or public consultation can depend on people trusting the evidence in front of them. If the source is unclear, the conversation often shifts away from the decision itself and towards whether the evidence can be believed.
VU.CITY's role is to make evidence easier to understand, scrutinise and share. That means bringing data into one spatial environment, but also making the context around that data visible: source, currency, ownership, definitions and confidence. Disagreement will always be part of good decision-making. A common, traceable evidence base makes those conversations more productive. In a low-trust environment, a shared and transparent evidence base is not a nice-to-have. It is the starting point for productive decisions. Provenance is what turns a visualisation from something persuasive into something accountable.
This is also where the next generation of spatial tools becomes interesting. If users can ask questions of geospatial data directly, filter it, explore it and understand what sits behind it, then data can become easier to explore while keeping the rigour people need to trust it.
We've been developing an early proof of concept for a conversational geospatial assistant that allows users to explore spatial datasets through natural language. Rather than requiring specialist GIS knowledge, users could ask a question in plain English: for example, "show buildings taller than 20 metres in this area that have been approved in the past 2 years" or "show all schools within a 500m radius of this address". The assistant would then plan and run the relevant spatial query, and cross-query, against trusted datasets, before rendering the results as layers in the model.
GIS specialists and professional judgement remain important, but there's an opportunity to reduce the barrier for non-specialist users who need to understand what data represents, interrogate it more easily and bring relevant outputs into a spatial workflow. Data provenance remains central: answers need to be based on known sources, with outputs that can be checked, challenged and understood. Complex spatial data becomes more valuable when more people can question it, understand it and use it with confidence.
To find out more about the new VU.CITY and our AI Data Lab, click here.