CRCarlos Duarte Roa

Carlos Duarte Roa

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Carlos Duarte Roa has a total of 51 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Knowledge and Semantic Graph-guided Large Language Models for Automated Fault Detection and Diagnosis in HVAC Systems

    by Aoyu Zou, Tong Xiao, Carlos Duarte Roa, Stefano Schiavon

    HVAC systems in buildings play a major role in energy use and operational carbon emissions, making reliable fault detection and diagnostics important for their efficient operation and occupant safety. However, conventional methods often face scalability challenges because they require site-specific rules, large labeled datasets, or limitedly interpretable models. In this study, we present an automated fault detection and diagnostic workflow that integrates a large language model with semantic and knowledge graphs. The proposed framework uses a semantic metadata model to represent HVAC system topology, a directed causal graph to encode physics-expected cause-and-effect relationships, and precomputed statistical evidence to guide large language model-based diagnostic inference. Rather than reasoning directly from raw sensor data, the LLM generates fault hypotheses grounded in a structured system context and quantitative evidence. We validated the workflow in a real controlled environmental chamber using a single-zone variable-air-volume air-handling unit operating under an ASHRAE Guideline 36 air-side economizer sequence. We included three representative faults: an overridden supply fan variablespeed drive, a stuck outdoor air damper, and a drifted supply air temperature sensor. Results show that the proposed workflow can identify the primary faults by interpreting cross-component interactions and fault propagation patterns. By grounding LLM reasoning in system ontology, expert knowledge, and statistical evidence, this approach improves diagnostic transparency, reduces hallucination risk, and provides a scalable pathway for AI-assisted HVAC fault diagnosis.

  • Open Access

    Room surface convective heat transfer with ceiling fans and its effect on radiant cooling systems

    by Carlos Duarte Roa, Paul Raftery, Sangjoon Lee, Aslihan Senel Solmaz

    1. Radiant heating and cooling systems are increasingly adopted as a strategy to reduce building energy use and greenhouse gas emissions compared to conventional all-air systems (Rhee and Kim 2015)...


  • Open Access

    Open-Tool Frameworks for Cross-Platform Indoor Monitoring and Optimized Air Cleaning Strategie

    by Federico Dallo, Lorenzo Tenti, Alessandro Palo, Thomas Parkinson, Carlos Duarte

    Indoor environments account for most human exposure to air pollution, yet indoor air quality (IAQ) monitoring and control remain fragmented across devices, platforms, and proprietary building automation systems. Commercial IAQ monitors and smart thermostats are widely available, but they typically operate in closed ecosystems with limited interoperability. In parallel, open-source communities have demonstrated the potential of low-cost sensing networks, yet these solutions rarely connect to building-level control systems capable of simultaneously reducing pollutant exposure and energy use. To address this gap, we present an open, interoperable framework that integrates open-source and commercial technologies for IAQ monitoring, data management, and automated building control[1]. The framework, developed within the EU-funded healthRiskADAPT project, is built on an open, production-ready IoT infrastructure for indoor environments. At the edge, low-cost sensor nodes collect and transmit environmental data. A web-based interface allows users to register locations, nodes, and sensors, and provides near-real-time visualization, historical analytics, and an interactive map of the sensor network. Beyond monitoring, the framework enables direct integration with commercial control devices such as smart thermostats, smart plugs, and filtration systems. This interoperability supports data-driven control strategies, including increasing ventilation during indoor pollution events, activating filtration during periods of poor outdoor air quality, and dynamically adjusting HVAC operation to balance comfort, energy use, and exposure reduction. By combining continuous mass-balance modeling[2] with real-time sensor data, the system will deliver actionable indoor-outdoor (I/O) ratios and exposure indicators. These outputs could drive automated responses but also support informed user behavior, such as choosing higher-efficiency filters during high-pollution episodes, using kitchen exhaust during cooking, or understanding the trade-offs between energy costs and health risks. In this way, the platform functions not only as a control system but also as an educational and decision-support tool for occupants and building managers. This presentation demonstrates how open-source hardware, open APIs, and modular integration pathways can create a flexible, transparent, and scalable ecosystem for IAQ management. The framework supports diverse use cases, homes, schools, workplaces, and research settings, while offering a roadmap toward energy-efficient, healthier indoor environments driven by interoperable technologies rather than isolated products.[1] https://particularmatter.org[2] Dallo, Federico, Thomas Parkinson, Carlos Duarte, Stefano Schiavon, Chai Yoon Um, Mark P. Modera, Paul Raftery, Carlo Barbante, and Brett C. Singer. "Using smart thermostats to reduce indoor exposure to wildfire fine particulate matter (PM2. 5)." Indoor Environments 2, no. 2 (2025): 100088.

  • Open Access

    Safe indoor temperature limits shape overheating resilience and cooling demand in California homes

    by Carlos Duarte Roa, Yan Wang, Harry Jiang, Charlie Huizenga, Aoyu Zou, Tong Xiao, Guoquan Lv, Tobias Kramer, Stefano Schiavon, Paul Raftery, et al.

    As climate change intensifies, the frequency and severity of extreme heat events are increasing, placing growing stress on building environments and energy systems. Using a physics-based residential building-stock framework, we conducted over 265,000 whole-building simulations (representing over 50,000 California homes across five climate scenarios) to quantify indoor overheating severity and the resulting increases in cooling demand. Indoor heat exposure was evaluated using the indoor overheating degree (IOD), which captures both the duration and magnitude of temperature exceedances above selected comfort limits. Results show that by the 2080s, up to ~ 55% of homes without cooling may exceed safe indoor limits, driving a 20–40% rise in statewide peak cooling electricity loads and potentially surpassing grid adequacy margins. Overheating risk moves from inland where AC is more common to coastal regions where it is not, increasing overall indoor heat exposure while narrowing regional disparities. By linking thermal limits with AC adoption and grid impacts, we quantify where overheating risks and cooling burdens intensify under current and future climates, providing evidence to support thermally safe and grid-aware housing policies.

  • Integrating symbolic neural networks with building physics: A study and proposal

    by Xia Chen, Guoquan Lv, Xinwei Zhuang, Carlos Duarte Roa, Stefano Schiavon, Philipp Geyer


  • Open Access

    Skewering the silos: using Brick to enable portable analytics, modeling and controls in buildings

    by Gabe Fierro, Johnson Controls International, Paul Raftery, Carlos Duarte Roa, Marco Pritoni, Michael Wetter, Anand Krishnan Prakash, Lazlo Paul, Erik Paulson, Therese Peffer

    Nearly all large commercial buildings have heating, ventilation and air conditioning (HVAC) systems, lighting systems, safety and other systems controlled by a computer—a dedicated server with a building energy management system (BMS). However, these BMSs are proprietary with each building’s assets (that is, fans, valves, pumps, and their setpoints) named and coded uniquely by the BMS vendor or engineer; building analytics and control algorithms are written specific to the assets and the building. Thus, any control updates or analytics to improve building performance—especially critical to reduce greenhouse emissions or improve load flexibility—are labor intensive and costly. The Brick schema was developed so the same analysis or control algorithms can work on a variety of buildings if each is digitally represented in a Brick data model. The goal of this project was to further the development of Brick to extend it beyond an academic project with demonstrated success in a small field study, to a practical choice for industrial and commercial stakeholders seeking to realize value from building data. To do this, we executed four objectives: (1) expand the Brick schema including its modeling capabilities and vocabulary, (2) develop tools for integrating Brick with existing digital technologies and representations in buildings,more » (3) develop an open-source analytics platform to facilitate use of Brick in delivering data value, and (4) demonstrate Brick-driven analytics and controls in real settings. « less