Impact Lab Wellington: Emergency Management
A one-day Impact Lab alongside Wellington City Council Emergency Management — applying Civic AI to improve the flow of information between communities and Council before and during an emergency.
Civic AI prototype built alongside Wellington City Council — not real emergency information. In an emergency call 111.
On Saturday 8 August 2026 we're bringing the Wellington tech community together with Wellington City Council Emergency Management for a one-day Impact Lab.
Teams will build working prototypes against real operational problem statements from the Council's emergency management team. The common theme: improving the flow and use of information between communities and Council before and during an event — from official warnings and street-level conditions to community reports and movement data.
Participants will only work with publicly available data. Every team's module plugs into one shared common operating picture: a live map and realtime feed of emergency "signals", so the room builds one product together.
Why your input matters
The best prototypes will be grounded in what people actually experienced. If you've lived through a south-coast storm, helped run a Community Emergency Hub, or know a public dataset the teams should be using — tell us below. Everything submitted is aggregated and handed to the teams (and their AI assistants) before the day.
Problem statements
Bring official warnings and local conditions into one clear community view
How might we give people a clear, location-specific picture of an emerging weather event by bringing together official warnings, Council information and trusted reports of local conditions?
South coast events are often forecast by MetService and communicated through official channels, but those sources don't always show what is happening at street or neighbourhood level — road conditions, waves crossing the road, surface flooding or access becoming unsafe. Residents monitor MetService, WCC, WREMO, news media and local Facebook groups without knowing which source to rely on or how the information fits together.
Desired outcome: People can quickly understand what is forecast, what is happening locally, and where to find authoritative advice.
Create a two-way information channel between communities and WCC
How might communities provide WCC with timely, structured information about local conditions, impacts and needs — before and during an emergency — and see that their information has been received?
The current flow is inconsistent and largely one-way: WCC sends information out, while reports from communities arrive through several unrelated channels and may not reach the people who can use them. A prototype could let residents, community groups or Community Emergency Hubs report an issue with location, time, issue type, description and an image — and see receipt acknowledged.
Desired outcome: WCC gains better local awareness, while communities have a clearer and more dependable route into Council.
Identify and verify emerging local impacts from public information
How might we use public online information to identify where emergency impacts may be emerging, while making the reliability and limitations of that information clear?
A prototype could collect relevant public posts, local news and community reports; identify likely locations and issue types; and show where several independent sources appear to describe the same event — presenting signals for an intelligence team to investigate, never social-media content as verified fact.
Desired outcome: WCC can detect possible impacts earlier and direct staff attention to matters needing confirmation.
Help emergency staff sort and prioritise incoming information
How might we help emergency staff rapidly sort incoming reports into information for awareness, information requiring verification, and information requiring action?
During an event, information arrives through phone calls, emails, forms, social media, news reports and partner agencies. Staff must identify where it relates to, whether it is new or duplicated, how reliable it may be, and whether it requires action. A prototype could extract location, time, issue and urgency; group similar reports; preserve source links; and present a queue for human review.
Desired outcome: Staff spend less time sorting information and more time checking significant reports and coordinating action.
Detect unusual changes in movement around the city
How might we identify and map sudden changes in pedestrian or vehicle movement that could indicate disruption, unsafe conditions, evacuation or loss of access?
A prototype could compare current movement with usual patterns and flag significant changes, correlating them with weather warnings, road closures or public reports. This could build on Pōneke Travel Insights, whose documented data limitations would need to stay visible in any emergency use.
Desired outcome: WCC receives another early indication of where an event may be affecting people, rather than relying only on individual reports.
Datasets
Publicly available data sources teams can build on during the lab.
Earthquake and geohazard feeds (CC BY) — GeoJSON and CAP alerts.
Road closures, delays and highway conditions.
Council open data portal, including hazard layers.
Regional environmental monitoring and hazard data.
Movement patterns, busy periods and change over time — note the documented data limitations.
Public news feeds for media reports of local conditions.