Award-winning AI triage pilot resolves over half of alarm calls without an operator, with zero missed emergencies

1​8 August 2026

Alcove, Yokeru and Enovation publish results of the UK's first live AI-enabled inbound triage pilot for an Alarm Receiving Centre, following the collaboration's win of the TSA 2026 Digital Innovation Award.

UK, 30 April 2026 - Alcove, Yokeru and Enovation have today published the results of a pilot evaluation showing that AI-enabled inbound triage can safely and effectively reduce avoidable demand on UK Alarm Receiving Centres (ARCs) - resolving more than half of inbound alarm calls without a human operator, while missing zero genuine emergencies.

The pilot, delivered with independent oversight from TEC Quality against the Quality Standards Framework (QSF), embedded a conversational AI voice agent, "Amy," into the live alarm-handling workflow at Alcove's TSA QSF-accredited ARC. Amy is built on Yokeru's AI care platform and operates through Enovation's UMOx connected care platform, which manages call routing, recording and audit trail. The collaboration was named winner of the TSA 2026 Digital Innovation Award.

RESULTS AT A GLANCE (AS OF 13 MARCH 2026)

• 446 calls presented to the AI triage agent
• 245 calls (54.9%) resolved by AI without operator involvement
• 100% of calls handled correctly
• Zero missed emergencies
• 300 live service users now on the AI triage pathway
• 22% increase in proactive, preventative referrals as freed-up operator time was reinvested in trend analysis and frequent-caller review

Of the 201 calls (45.1%) escalated to a human operator, every escalation was for an appropriate reason - including genuine emergencies, no speech detected, inability to confidently rule out a false alarm, carer requests, and equipment test calls.

ADDRESSING A SECTOR-WIDE PROBLEM

More than half of all inbound calls to ARCs are false alarms or test calls that require no operator intervention, and false alarms alone can account for around 40% of operator time - time taken away from genuine emergencies and proactive care. As connected devices generate more data, ARC workloads have more than tripled in recent years without a corresponding increase in funding, leaving some centres struggling to keep pace.

When an alarm is triggered by a wearable device - such as a pendant, falls detector, GPS tracker or push-button device - the call is routed to Amy via UM Ox. Amy conversationally and safely establishes whether the alert is a genuine emergency using structured questioning and double confirmation. If the caller confirms they are safe, Amy closes the call and logs the outcome; if there is any uncertainty, distress, no response, or a genuine emergency, the call is escalated to a human operator within seconds. High-risk alert types - including smoke and fire alarms, medical monitoring alerts, and service users with safeguarding concerns - are excluded from AI triage entirely and always routed directly to an operator.

BUILT FOR SAFETY FIRST

The system was designed with no single point of failure: if the AI cannot connect, the call disconnects, or anything unexpected occurs, the call is automatically routed to a human operator. The pilot followed an eight-week staff-led testing phase before any service user involvement, with all outcomes logged, reviewed and formally signed off, and eligibility for AI triage continues to be reviewed dynamically as service users' needs and risk profiles change.

"Continue refining and testing until 100% accuracy is achieved prior to implementation."- Claire Aldridge, Head of Client Development & Clinical Oversight, Alcove, on the threshold set for the pilot before go-live

"54.9% of activations resolved without an operator, zero missed emergencies - that tells us the safeguards hold up at scale. The real outcome is what it.frees up: teams back on genuine emergencies and proactive care instead of triaging false alarms. That's the direction the whole sector needs to move in." - Monty Alexander, CEO, Yokeru

SCALING ACROSS THE SECTOR

Projected across Alcove's full ARC customer base, the approach could save an estimated 17,000+ operator hours per year - equivalent to over £300,000 in workforce efficiency savings annually - which the partners say would be reinvested into faster emergency response and more proactive, preventative support, without additional staffing or funding.

The pilot is now moving from evaluation to standard operating procedure, with AI triage set to become the default approach for handling eligible alarm activations across Alcove's full ARC. Alcove, Yokeru and Enovation will continue to work with TEC Quality and the TSA Sector Risk and Innovation Group (SRIG) to feed learnings from the pilot into the future development of the QSF and wider sector standards for safe AI adoption.

R​ead the published report