'HOME'CAST: Shaping the Built Environment through Data-Driven Innovation | Keynote Address

Item: 'HOME'CAST: Shaping the Built Environment through Data-Driven Innovation | Keynote Address
Rating: 4
Author: PostSingularity Institute

by Harvard GSD

Applied Sciences & Engineering Arts & Architecture AMCArchitectural structure and design AMCREnvironmentally-friendly and sustainable architecture and design

🎙 Harvard GSD 👥 124K 📅 October 10, 2025 ⏱ 48 min 👁 253 📄 expert opinion 🧭 2026-08-16

Available in: English (current) Français

Keywords

building energy modelingmachine learningHVAC controlindoor air qualityurban context

Summary

The keynote address by Holly Samuelson at the ‘HOME’CAST conference focuses on the intersection of building design, technology, and climate. Samuelson, an associate professor at MIT, shares ten research tips for working in this field. She emphasizes the importance of using simple models when appropriate, combining models to leverage their strengths, and making educated guesses for unknown variables. She discusses the application of reinforcement learning to building controls, borrowing methods from other fields, and even from nature, such as mimicking a dog’s sniffing pattern to improve chemical sensors for indoor air quality. The talk highlights the need for better tools for early-stage design and the potential of data-driven approaches to address challenges like energy efficiency, occupant comfort, and health. Samuelson concludes with a call for innovation and collaboration to accelerate progress in creating resilient and sustainable buildings.

137 words

Critical Evaluation

Value of the Information & Strength of the Argument

The talk provides valuable insights from a practitioner and researcher with extensive experience. The argumentation is solid, grounded in specific examples from her own research and that of her lab. She effectively uses case studies, such as the calibration of a building model that saved $130,000 per year, to illustrate the practical value of her points. The advice is actionable and well-reasoned, encouraging a pragmatic approach to modeling and research. However, the talk is more of a collection of tips than a systematic argument, and some points could benefit from deeper exploration.

Scientific Rigor, Source Quality, Title Accuracy

The speaker is highly credible, with over 50 peer-reviewed publications and recognition in the field. She references her own work and that of colleagues, but does not provide specific citations during the talk. The title accurately reflects the content, focusing on data-driven innovation in the built environment. The talk is well-structured and the content aligns with the conference theme. The lack of explicit citations is a minor weakness, but the speaker’s authority and the practical nature of the talk mitigate this.

188 words

Title / Content Match

The title accurately reflects the content: a keynote address on data-driven innovation in the built environment, focusing on AI, predictive analytics, and climate modeling for housing.

Quality & Reliability

8/10

The keynote is delivered by a recognized expert (Holly Samuelson) with a strong publication record, and it presents practical research insights grounded in peer-reviewed work. The content is well-structured and evidence-based, though it is primarily an opinion/experience-based talk rather than a formal literature review or original study.

Chapters

Contribution & Novelties

The talk offers a practitioner’s perspective on integrating data-driven methods into building design, with a focus on practical, actionable tips. It highlights the importance of using simple models when appropriate, combining models, and borrowing techniques from other fields. The discussion of using reinforcement learning for HVAC control and the novel approach to improving chemical sensors by mimicking a dog’s sniffing pattern are particularly innovative. The talk also emphasizes the need for better tools for early-stage design and the potential of AI to address challenges in the built environment.

Pour aller plus loin :

Reinforcement learning — Relevant to the discussion of using RL for building controls.
Building energy simulation — Provides background on the modeling techniques discussed.
Indoor air quality — Context for the sensor development and health impacts.

129 words

Radar Profile

The radar profile shows high scores in information quantity, quality, and reliability, with a slightly lower technical level. This indicates a well-informed and credible talk that is accessible to a broad audience, though it may not delve into the most advanced technical details.

Reliability 8/10