Image by Rawpixel / Freepik www. refrigerationandaircon. co. za RACA Journal I June 2025 15
Business
When prompted to create a unit test for a function calculating the average of an array, the AI swiftly generated a test method, even offering suggestions for assertion functions and auto-completing code snippets.
Dall’ Occo showcased STone Copilot ' s remarkable ability to generate code directly from visual diagrams. By simply copying an image of a basic control logic diagram into STone, he prompted the AI to " create a function block that behaves like it is shown in the image ". The near-instantaneous generation of functional code from a visual representation captured the potential of this feature to bridge the gap between conceptual diagrams and deployable code.
By automating mundane yet critical tasks like commenting and unit testing, and by intelligently generating and optimising code, STone promises to free up developers to focus on more complex and innovative aspects of their work. The potential time savings of 30 % in comment generation and 20 % in unit test generation, as highlighted in the webinar, underscores the tangible value proposition of this cutting-edge ecosystem. As the HVAC & R sector increasingly embraces digitalisation and smart technologies, STone IDE is being positioned as an indispensable asset for developing high-quality, efficient and secure software solutions.
BRIDGING LANGUAGE BARRIERS AND SIMPLIFYING COMPLEXITIES IN HVAC & R SOFTWARE DEVELOPMENT
Dall’ Occo showcased its remarkable ability to understand and respond in other languages, effectively breaking communication barriers for international development teams. When Putzolu inquired
The code optimisation feature addresses the challenges of resource-constrained hardware. about interacting with STone Copilot in Italian, Dall’ Occo seamlessly switched to his mother tongue, asking the AI to " create a function block that counts from 10 to 45 every five seconds ". The AI responded promptly in Italian providing both the code and a clear explanation of its functionality.
Following this demonstration, Putzolu requested a concise summary of the benefits showcased. Dall’ Occo recapped the key advantages of integrating AI within STone. " This frees developers to concentrate on core algorithms and the intricacies of HVAC unit temperature control. Furthermore, the code optimisation feature addresses the challenges of resource-constrained hardware, while the unit test generation and code condition capabilities can save another 20 % of testing time, allowing developers to focus on the thermodynamic and hybrid aspects of their work.”
Putzolu concluded, " Generative AI in STone brings benefits in terms of productivity, code quality, continuous learning, and team collaboration. So at the end, it speeds up the software development department by four times." This emphasises the significant acceleration in the software development lifecycle. RACA
Image by Rawpixel / Freepik www. refrigerationandaircon. co. za RACA Journal I June 2025 15