Improve low-carbon heating and comfort outcomes with human-response intelligence

Low-carbon heating and HVAC systems can optimise technical performance, but realised outcomes also depend on occupant comfort and response.

EluxEvo is a UK deep-tech SME developing Human Acceptance Intelligence (HAI). Our AI/ML approach combines connected-device and environmental data with human interaction and behavioural signals, learning patterns of comfort, participation and response. HAI is being developed to make these human factors measurable alongside conventional technical and energy data.

We are seeking collaboration with heat-pump and HVAC manufacturers, smart heating and comfort-control providers, and connected low-carbon heating technology developers with a real-world challenge where occupant response materially affects intended outcomes.

A collaboration could explore whether HAI can provide a complementary human-response intelligence layer alongside existing sensing and control, helping partners better understand the relationship between thermal interventions, occupant response and realised comfort or energy outcomes.

What EluxEvo brings

EluxEvo has secured and led Innovate UK grant-funded R&D and participates in Digital Catapult’s Applied AI Growth Programme. Our HAI capability is at TRL 5, progressing towards TRL 6, with development and testing through smart-home environments and pilots.

What we are looking for

We are particularly interested in a technology manufacturer or controls provider with an existing connected heating or HVAC capability, a defined deployment challenge and access to relevant system data and users. The project is at design stage, and we want to co-design the use case, integration and validation approach around that challenge.

BridgeAI Deployment Catalyst

We see AI for Operational Systems as the most likely initial fit, where human-response intelligence could complement an existing AI-enabled heating, comfort or control system. Other Catalyst pathways could be considered where the partner's specific deployment challenge supports them.