What does \u201cAI in building design\u201d actually mean, and what does it look like when it works in practice?<\/p>\n\n\n\n
In our new episode of the Electrical Building Design Show, we sat down with Kevin Lawson, PE, founder of Ripple, to discuss how automation and AI are being applied to HVAC design, and what those lessons mean for the broader MEP industry.<\/p>\n\n\n\n
Kevin brings a unique perspective to the conversation. After a decade designing and commissioning HVAC and plumbing systems, he transitioned from consulting engineering into software development, founding Ripple with a clear goal: automate HVAC design directly inside Revit.<\/p>\n\n\n\n
One of the central themes of the discussion is the role HVAC plays as a critical dependency for nearly every other discipline. Electrical loads, structural coordination, mechanical room sizing, and even architectural decisions often hinge on HVAC being designed first and done so accurately.<\/p>\n\n\n\n
By accelerating HVAC design, Kevin argues, the entire project team benefits. Faster access to reliable loads, equipment selections, and system data can dramatically reduce downstream friction for electrical, structural, and architectural teams.<\/p>\n\n\n\n
Rather than focusing on large language models or black-box neural networks, Kevin frames AI more practically: If a system takes an input and produces a human-like design output, it qualifies as AI, regardless of how it gets there.<\/strong><\/p>\n\n\n\n Ripple\u2019s approach is procedural, not probabilistic. It follows the same step-by-step logic an engineer would use: load calculations, ventilation requirements, equipment selection, and system layout, automated at scale inside Revit. The result is an HVAC model that looks and behaves as if it were designed by an engineer but produced in a fraction of the time.<\/p>\n\n\n\n A key question addressed in the episode is whether architectural Revit models are \u201cgood enough\u201d to support automated engineering workflows. The answer, backed by real-world validation, may surprise you.<\/p>\n\n\n\n Kevin shares results from a hospital study where automated load calculations, run directly from an architect\u2019s model, matched measured building performance within 1%. Beyond accuracy, the conversation highlights how architects already curate much of the data engineers need, from window geometry to room definitions, making automation far more viable than many assume.<\/p>\n\n\n\n The discussion also explores how automation compounds in value. Once systems are laid out automatically, additional insights, pressure loss, fan sizing, material quantities, and cost data become far easier to generate. What starts as time savings in layout quickly evolves into richer, more reliable design data across disciplines.<\/p>\n\n\n\n Importantly, this isn\u2019t about futuristic promises. Much of what\u2019s described feels like the fulfillment of BIM\u2019s original promise: connected, data-rich models that support real engineering decisions.<\/p>\n\n\n\n Kevin offers a candid take on the growing number of AI demos in the industry, emphasizing skepticism without tangible design outputs. For engineers, proof matters. If a tool claims to automate design, the expectation should be simple: show the model.<\/p>\n\n\n\n This practical mindset carries through the broader conversation about platforms, workflows, and why staying embedded in Revit remains critical for real-world adoption.<\/p>\n\n\n\n The episode closes with a forward-looking discussion on productivity, technology adoption, and business models in AEC. Automation isn\u2019t just a technical shift; it has implications for how firms bill, how value is measured, and how engineering services are delivered.<\/p>\n\n\n\n Kevin challenges firms to think beyond billable hours and toward output-driven value, suggesting that automation will ultimately force a rethink of long-standing industry norms.<\/p>\n\n\n\n Watch the full video conversation between Dave and Kevin below to learn more.<\/p>\n\n\n\nLeveraging Architectural Models for Real Engineering Work<\/strong><\/h2>\n\n\n\n
Automation as a Snowball Effect<\/strong><\/h2>\n\n\n\n
Cutting Through AI Hype in AEC<\/strong><\/h2>\n\n\n\n
What This Means for Engineering Firms<\/strong><\/h2>\n\n\n\n
\n\n\n\n