All work

NHS England2 min read · 5 sections

Fault reporting in under a minute, for frontline NHS staff.

A mobile-first service that replaced a phone-based fault reporting process with a photo-first flow built for busy wards, shared devices and patchy Wi-Fi. Delivered as UX Designer alongside a Lead UX Designer, via TPXimpact.

NHS · Public sector · GDS

Role
UX Designer, working with a Lead UX Designer
Duration
6 months
Domain
NHS estates and facilities
Standards
GDS and NHS service standards
Start screen with emergency guidance to call the helpdesk
Fault location step asking whether the reporter knows the location ID
Report sent confirmation promising an update within two working days
average time to report a fault, down from minutes
<60s
more faults reported in the first quarter
340%
NHS trusts in four weeks of fieldwork
3

At a glance

Problem
Frontline staff reported faults by phone, at around three minutes a report, so many faults simply went unreported.
My role
UX Designer alongside a Lead UX Designer, via TPXimpact, including four weeks of fieldwork across three NHS trusts.
Key decision
Minimum viable input: location by QR code, photo-first capture, offline-first sync and status updates that close the loop.
Outcome
Average report time under 60 seconds, and 340% more faults reported in the first quarter.
On this pageThe stakes
  1. The stakes
  2. Research
  3. Design approach
  4. Impact
  5. Reflection

The stakes

NHS trusts manage billions in physical assets, and when equipment or facilities fail, patient care feels it. Frontline staff at major trusts were reporting faults by phone, taking around three minutes per report.

That sounds reasonable until you multiply it across hundreds of reports a day, mid-shift, on shared devices with poor connectivity. It was enough friction that many faults simply went unreported, and unreported faults became unresolved ones, compounding into safety risks and reactive maintenance costs.

Research

I ran four weeks of fieldwork across three NHS trusts: contextual inquiry on wards, in-depth interviews with facilities managers, and surveys of nurses, porters, facilities managers and administrators. Three findings shaped everything:

Reporting felt pointless
Staff never saw what happened after they reported. Reports seemed to disappear, which eroded trust and discouraged the next one.
The environment was hostile to interruption
Shared or ageing phones, intermittent Wi-Fi, a busy ward round. Any snag, from a mandatory dropdown to a dropped signal, meant the report was abandoned.
Classification was a bottleneck
Staff couldn't tell whether a fault was already reported, so duplicates wasted everyone's time.

Design approach

Location by QR code, not dropdown
Staff scan a code on the equipment or room, removing a multi-step lookup that ate much of the reporting time. I explored GPS, but it couldn't give room-level accuracy inside hospital buildings.
Photo-first capture
A photo replaced several descriptive fields: less effort for the reporter, richer context for maintenance.
Offline-first
Reports queue locally and sync in the background, so unreliable NHS Wi-Fi never blocks anyone.
Closing the loop
Status updates show reporters their fault was acknowledged, assigned and resolved, rebuilding the trust the old process had broken.
  • Fault location step asking whether the reporter knows the location ID
    Location ID first
  • Guidance on how to find a location ID with example photos
    Help finding it
  • Guidance on how to find an asset tag with example photos
    Asset tags explained
  • Report sent confirmation
    Loop closed

Impact

increase in fault reports in the first quarter
340%
average report time, down from around three minutes
<60s
reduction in reporting time
67%

Together these tell one story: reducing friction didn't just make reporting faster, it made reporting happen. The old process wasn't only slow, it was suppressing reports altogether.

Reflection

In healthcare, friction isn't just a usability problem, it's a patient safety problem. The counterintuitive lesson was that the biggest gain came from making reporting feel worth doing. Closing the feedback loop mattered as much as cutting form fields.

What I'd do differently
Push for integration with the trust's CAFM system much earlier, rather than treating it as phase two.
What I learnt
Prototype inside the constraints first. Designing the ideal and then compromising costs more.

Working on something this complex?

Lead and Principal roles, contract or permanent, or a focused engagement through Holistic Design AI. Tell me about it.