CAREL publishes white paper on generative AI use in HVACR

BRUGINE, Italy, 30 June 2026: CAREL said it published a white paper on the use of generative artificial intelligence in the HVACR sector, entitled “Gen AI in HVAC/R: making it more reliable. From prompt engineering to context engineering”.

Making the announcement through a June 30 Press Release, CAREL said Giandomenico Lombello, Senior Advisor of the CAREL Group and member of the Group’s AI Committee, wrote the white paper. CAREL said the document starts from the premise that there is no single prompt capable of making generative AI reliable on its own, and that in the HVACR sector, where decisions may have technical, operational, economic and safety implications, the quality of outputs depends on the quality of the context provided to the model.

CAREL said this is the basis for the white paper’s shift from prompt engineering to context engineering, which it described as not only asking better questions, but building the information conditions that allow the model to produce responses consistent with the real case. CAREL added that the document identifies risks beyond hallucinations, including plausible but insufficiently anchored answers that are formally correct yet inconsistent with the physical, plant or operating conditions at hand.

CAREL said the white paper frames generative AI as a form of augmented intelligence, a tool that does not replace expertise, judgement and experience but works alongside them. CAREL added that the document identifies security, data handling and compliance as prerequisites when working with service reports, plant logs, customer specifications, drawings, tender documents or technical documentation.

CAREL said the white paper covers 10 operational applications, including the Zoom Out/Zoom In method, multimodality, AI Projects, grounding and retrieval-augmented generation (RAG), role-based review and Deep Search.