Soter tested in real-world conditions
Whaikaha Access Activator Trial
A trial of Soter in 8 Dunedin households was funded by an Access Activator grant from Whaikaha – Ministry of Disabled People. The trial ran from early May to the end of July 2026.
The trial tested whether an accessibility-first doorstep assistant could make routine front-door interactions safer, calmer, and more manageable. The focus was on practical use in real homes rather than a laboratory demonstration.
It asked whether Soter could help households understand who was at the door and why, reduce stress or urgency, improve control, support different accessibility needs, and identify what needed to change before wider deployment.
The trial aims
- Test whether the visitor-first interaction model works in real household settings.
- Assess whether Soter helps participants understand who is at the door and why.
- Explore whether Soter reduces uncertainty, stress, effort, or reliance on others.
- Identify which accessibility features and communication modes are most important.
- Understand whether participants, whānau, and carers find Soter acceptable and trustworthy.
- Identify technical, practical, privacy, and support issues that must be addressed before wider rollout.
- Use co-design feedback to shape the next version of Soter.
Who took part
The pilot includes eight households in Ōtepoti Dunedin. The households were selected to provide real-world diversity across access needs, household situations and front-door contexts, while keeping the trial small enough to support intensive onboarding, troubleshooting and co-design.
The trial in numbers
Across eight Dunedin households, the trial captured a substantial volume and variety of real-world doorstep activity, providing a strong evidence base for understanding how Soter performed in everyday use.
What participants said
From the households in the Dunedin pilot
Some trial findings
After relevant visits, people were surveyed on whether Soter helped and why. Among completed surveys, participants answered yes 45.4%, unsure 10.2%, and no 44.4%. In cases where Soter was found helpful, ‘knowing who was there’ was selected 98% of the time; ‘knowing why’ 81.6% of the time; ‘communication’, 46.9%; ‘recording of the visit’, 38.8%; not needing to hurry, 18.4%; and notifying someone else, 10.2%.
What changed because of the pilot
The pilot was designed not only to test Soter, but to improve it through direct feedback from participants and households. Participants’ access needs, household routines, visitor patterns and feedback have informed changes to the product, configuration, support process and future roadmap.
Co-design inputs included baseline interviews, onboarding discussions, weekly check-ins, mid-pilot interviews, survey comments, technical support interactions and direct feature requests. Examples of features implemented during the trial due to participant feedback include:
- A dedicated indoor unit introduced during the pilot.
- A screen-reader-friendly occupant app developed with a blind participant.
- Door-left-open alerts.
- Appointment and expected-visitor checks.
- PIN workflows for trusted visitors.
