Bharathan Balaji
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Bharathan Balaji has a total of 69 co-authors. Here are the top co-authors by number of shared publications:
- Dezhi Hong
Department of Computer Science and Engineering, Halicioglu Data Science Institute, University of California, La Jolla, CA, USA
3 shared publications - Mani Srivastava
University of California, Los Angeles, Los Angeles, CA, USA
3 shared publications - Rajesh K. Gupta
University of California, San Diego, La Jolla, CA, USA
3 shared publications - Geoffrey Guest
Amazon, USA
2 shared publications - Hsin-Yu Liu
University of California, San Diego, La Jolla, CA, USA
2 shared publications - Venkata Sai Gargeya Vunnava
Amazon, USA
2 shared publications
Latest Publications
- Open Access
- 22 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 584-596
Extracting Product Carbon Footprint in PDF Documents using Question Answering Framework
- 22 June 2026
- Published by Association for Computing Machinery (ACM)
- p. 584-596
Open AccessExtracting Product Carbon Footprint in PDF Documents using Question Answering Framework
- Open Access
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 147-161
Concurrency-Informed Orchestration for Serverless Functions
- 30 March 2025
- Published by Association for Computing Machinery (ACM)
- p. 147-161
Open AccessConcurrency-Informed Orchestration for Serverless Functions
- Top Cited Papers
- 21 March 2025
- Published by American Chemical Society (ACS) in Environmental Science & Technology
- Vol. 59 (18) , 9113-9122
Emission Factor Recommendation for Life Cycle Assessments with Generative AI
Accurately quantifying greenhouse gas (GHG) emissions is crucial for organizations to measure and mitigate their environmental impact. Life cycle assessment (LCA) estimates the environmental impacts throughout a product's entire lifecycle, from raw material extraction to end-of-life. Measuring the emissions outside a product owner's control is challenging, and practitioners rely on emission factors (EFs)─estimations of GHG emissions per unit of activity─to model and estimate indirect impacts. However, the current practice of manually selecting appropriate EFs from databases is time-consuming and error-prone and requires expertise. We present an AI-assisted method leveraging natural language processing and machine learning to automatically recommend EFs with human-interpretable justifications. Our algorithm can assist experts by providing a ranked list of EFs or operating in a fully automated manner, where the top recommendation is selected as final. Benchmarks across multiple real-world data sets show our method recommends the correct EF with an average precision of 86.9% in the fully automated case and shows the correct EF in the top 10 recommendations with an average precision of 93.1%. By streamlining EF selection, our approach enables scalable and accurate quantification of GHG emissions, supporting organizations' sustainability initiatives and progress toward net-zero emissions targets across industries.
- 21 March 2025
- Published by American Chemical Society (ACS) in Environmental Science & Technology
- Vol. 59 (18) , 9113-9122
Top Cited PapersEmission Factor Recommendation for Life Cycle Assessments with Generative AI
Accurately quantifying greenhouse gas (GHG) emissions is crucial for organizations to measure and mitigate their environmental impact. Life cycle assessment (LCA) estimates the environmental impacts throughout a product's entire lifecycle, from raw material extraction to end-of-life. Measuring the emissions outside a product owner's control is challenging, and practitioners rely on emission factors (EFs)─estimations of GHG emissions per unit of activity─to model and estimate indirect impacts. However, the current practice of manually selecting appropriate EFs from databases is time-consuming and error-prone and requires expertise. We present an AI-assisted method leveraging natural language processing and machine learning to automatically recommend EFs with human-interpretable justifications. Our algorithm can assist experts by providing a ranked list of EFs or operating in a fully automated manner, where the top recommendation is selected as final. Benchmarks across multiple real-world data sets show our method recommends the correct EF with an average precision of 86.9% in the fully automated case and shows the correct EF in the top 10 recommendations with an average precision of 93.1%. By streamlining EF selection, our approach enables scalable and accurate quantification of GHG emissions, supporting organizations' sustainability initiatives and progress toward net-zero emissions targets across industries.
- Open Access
- 12 March 2025
- Published by Elsevier in International Journal of Applied Earth Observation and Geoinformation
- Vol. 138
FUELVISION: A multimodal data fusion and multimodel ensemble algorithm for wildfire fuels mapping
- 12 March 2025
- Published by Elsevier in International Journal of Applied Earth Observation and Geoinformation
- Vol. 138
Open AccessFUELVISION: A multimodal data fusion and multimodel ensemble algorithm for wildfire fuels mapping
- Open Access
- 26 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 125
Towards a Performance-Driven Device-Edge-Cloud Relationship
Real-time cyber-physical systems (CPS) rely on Perception-Cognition-Actuation (PCA) pipelines to enable autonomous observation, decisionmaking, and action execution. Closed-loop PCA systems utilize feedback-driven control to iteratively adapt actions in response to real-time environmental changes whereas open-loop PCA systems execute single actions without iterative feedback. The overall performance of these systems is inherently tied to the models selected for each pipeline component. Recent advancements in neural networks, particularly for perception tasks, have substantially enhanced CPS capabilities but have introduced significant complexity into the PCA pipeline. While traditional research [1] often evaluates perception models in static, controlled settings, it fails to account for the cascading latency and accuracy trade-offs that manifest across interconnected PCA modules in dynamic, real-time applications. Additionally, the proliferation of distributed device-edge-cloud architectures [2] has expanded computational possibilities but introduced new challenges in balancing latency and accuracy with resource constraints. The holistic impact of model selection, deployment platforms, and network conditions on application performance in real-time scenarios remains under-explored.
- 26 February 2025
- Published by Association for Computing Machinery (ACM)
- p. 125
Open AccessTowards a Performance-Driven Device-Edge-Cloud Relationship
Real-time cyber-physical systems (CPS) rely on Perception-Cognition-Actuation (PCA) pipelines to enable autonomous observation, decisionmaking, and action execution. Closed-loop PCA systems utilize feedback-driven control to iteratively adapt actions in response to real-time environmental changes whereas open-loop PCA systems execute single actions without iterative feedback. The overall performance of these systems is inherently tied to the models selected for each pipeline component. Recent advancements in neural networks, particularly for perception tasks, have substantially enhanced CPS capabilities but have introduced significant complexity into the PCA pipeline. While traditional research [1] often evaluates perception models in static, controlled settings, it fails to account for the cascading latency and accuracy trade-offs that manifest across interconnected PCA modules in dynamic, real-time applications. Additionally, the proliferation of distributed device-edge-cloud architectures [2] has expanded computational possibilities but introduced new challenges in balancing latency and accuracy with resource constraints. The holistic impact of model selection, deployment platforms, and network conditions on application performance in real-time scenarios remains under-explored.
- Open Access
- 29 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Adaptive Policy Regularization for Offline-to-Online Reinforcement Learning in HVAC Control
Reinforcement learning (RL)-based control methods have been extensively studied to improve building heating, ventilation, and air conditioning (HVAC) efficiency. Data-driven approaches demonstrate better transferability and scalability, making them useful in real-world applications. Most prior works focus on online learning requiring simulators or models of environment dynamics. However, transferring thermal simulators between environments is inefficient in practice. We build on recent works that employ offline training on static datasets from unknown policies. Pure offline RL is constrained by the replay buffer's distribution, we propose using offline-to-online RL to enhance pre-trained offline models through online adaptation to distribution shifts. We show that direct online fine-tuning deteriorates performance on offline policies. To address this, we propose automatically tuning the actor's regularization during training to optimize the exploration-exploitation tradeoff. Specifically, we leverage simple moving averages of mean Q-values sampled throughout training. Simulation experiments demonstrate our method outperforms state-of-the-art approaches under various conditions, improving performance by 32.9% and enhancing pre-trained models' capabilities online.
- 29 October 2024
- Published by Association for Computing Machinery (ACM)
- p. 1-10
Open AccessAdaptive Policy Regularization for Offline-to-Online Reinforcement Learning in HVAC Control
Reinforcement learning (RL)-based control methods have been extensively studied to improve building heating, ventilation, and air conditioning (HVAC) efficiency. Data-driven approaches demonstrate better transferability and scalability, making them useful in real-world applications. Most prior works focus on online learning requiring simulators or models of environment dynamics. However, transferring thermal simulators between environments is inefficient in practice. We build on recent works that employ offline training on static datasets from unknown policies. Pure offline RL is constrained by the replay buffer's distribution, we propose using offline-to-online RL to enhance pre-trained offline models through online adaptation to distribution shifts. We show that direct online fine-tuning deteriorates performance on offline policies. To address this, we propose automatically tuning the actor's regularization during training to optimize the exploration-exploitation tradeoff. Specifically, we leverage simple moving averages of mean Q-values sampled throughout training. Simulation experiments demonstrate our method outperforms state-of-the-art approaches under various conditions, improving performance by 32.9% and enhancing pre-trained models' capabilities online.
- Open Access
- 6 December 2023
- Published by Association for Computing Machinery (ACM) in ACM Journal on Computing and Sustainable Societies
- Vol. 1 (2) , 1-23
Flamingo: Environmental Impact Factor Matching for Life Cycle Assessment with Zero-shot Machine Learning
Consumer products contribute to more than 75% of global greenhouse gas (GHG) emissions, primarily through indirect contributions from the supply chain. Measurement of GHG emissions associated with products is a crucial step toward quantifying the impact of GHG emission abatement actions. Life cycle assessment (LCA), the scientific discipline for measuring GHG emissions, estimates the environmental impact associated with each stage of a product from raw material extraction to its disposal. Scaling LCA to millions of products is challenging as it requires extensive manual analysis by domain experts. To avoid repetitive analysis, environmental impact factors (EIFs) of common materials and products are published for use by LCA experts. However, finding appropriate EIFs for even a single product under study can require hundreds of hours of manual work, especially for complex products. We present Flamingo, an algorithm that leverages natural language machine learning (ML) models to automatically identify an appropriate EIF given a text description. A key challenge in automation is that EIF databases are incomplete. Flamingo uses industry sector classification as an intermediate layer to identify when there are no good matches in the database. On a dataset of 664 products, our method achieves an EIF matching precision of 75%.
- 6 December 2023
- Published by Association for Computing Machinery (ACM) in ACM Journal on Computing and Sustainable Societies
- Vol. 1 (2) , 1-23
Open AccessFlamingo: Environmental Impact Factor Matching for Life Cycle Assessment with Zero-shot Machine Learning
Consumer products contribute to more than 75% of global greenhouse gas (GHG) emissions, primarily through indirect contributions from the supply chain. Measurement of GHG emissions associated with products is a crucial step toward quantifying the impact of GHG emission abatement actions. Life cycle assessment (LCA), the scientific discipline for measuring GHG emissions, estimates the environmental impact associated with each stage of a product from raw material extraction to its disposal. Scaling LCA to millions of products is challenging as it requires extensive manual analysis by domain experts. To avoid repetitive analysis, environmental impact factors (EIFs) of common materials and products are published for use by LCA experts. However, finding appropriate EIFs for even a single product under study can require hundreds of hours of manual work, especially for complex products. We present Flamingo, an algorithm that leverages natural language machine learning (ML) models to automatically identify an appropriate EIF given a text description. A key challenge in automation is that EIF databases are incomplete. Flamingo uses industry sector classification as an intermediate layer to identify when there are no good matches in the database. On a dataset of 664 products, our method achieves an EIF matching precision of 75%.
- Open Access
- 15 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 429-433
Quantifying the Decarbonization Potential of Flexible Load
The impact of human activity on the climate is a major global challenge that affects human well-being. Buildings are a major source of energy consumption and carbon emissions worldwide, especially in advanced economies such as the United States. As a result, making grids and buildings sustainable by reducing their carbon emissions is emerging as an important step toward societal decarbonization and improving overall human well-being. While prior work on demand response methods in power grids and buildings has targeted peak shaving and price arbitrage in response to price signals, it has not explicitly targeted carbon emission reductions. In this paper, we analyze the flexibility of building loads to quantify the upper limit on their potential to reduce carbon emissions, assuming perfect knowledge of future demand and carbon intensity. Our analysis leverages real-world demand patterns from 1000+ buildings and carbon-intensity traces from multiple regions. It shows that by manipulating the demand patterns of electric vehicles, heating, ventilation, and cooling (HVAC) systems, and battery storage, we can reduce carbon emissions by 26.93% on average and by 54.90% at maximum. Our work advances the understanding of sustainable infrastructure by highlighting the potential for infrastructure design and interventions to significantly reduce carbon footprints, benefiting human well-being.
- 15 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 429-433
Open AccessQuantifying the Decarbonization Potential of Flexible Load
The impact of human activity on the climate is a major global challenge that affects human well-being. Buildings are a major source of energy consumption and carbon emissions worldwide, especially in advanced economies such as the United States. As a result, making grids and buildings sustainable by reducing their carbon emissions is emerging as an important step toward societal decarbonization and improving overall human well-being. While prior work on demand response methods in power grids and buildings has targeted peak shaving and price arbitrage in response to price signals, it has not explicitly targeted carbon emission reductions. In this paper, we analyze the flexibility of building loads to quantify the upper limit on their potential to reduce carbon emissions, assuming perfect knowledge of future demand and carbon intensity. Our analysis leverages real-world demand patterns from 1000+ buildings and carbon-intensity traces from multiple regions. It shows that by manipulating the demand patterns of electric vehicles, heating, ventilation, and cooling (HVAC) systems, and battery storage, we can reduce carbon emissions by 26.93% on average and by 54.90% at maximum. Our work advances the understanding of sustainable infrastructure by highlighting the potential for infrastructure design and interventions to significantly reduce carbon footprints, benefiting human well-being.
- Open Access
- 15 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 366-369
Economizer Optimization with Reinforcement Learning: An Industry Perspective
Building operations contribute approximately 28% of global greenhouse gas emissions according to the International Energy Agency. With the increase in cooling demand due to rising global temperatures, the optimization of rooftop units (RTUs) in buildings becomes crucial for reducing emissions. We focus on the optimization of the economizer logic within RTUs, which balances the mix of indoor and outdoor air. By effectively utilizing outside air, RTUs can significantly decrease mechanical energy usage, leading to reduced energy costs and emissions. However, the current practice of economizer optimization relies on static guidelines set by ASHRAE, which approximates the dynamics of individual facilities. We introduce a reinforcement learning (RL) approach that adaptively controls the economizer based on the unique characteristics of individual facilities. We have deployed our solution in the real-world across a distributed building stock. We address the scaling challenges with our cloud-based RL deployment on 10K+ RTUs across 200+ sites.
- 15 November 2023
- Published by Association for Computing Machinery (ACM)
- p. 366-369
Open AccessEconomizer Optimization with Reinforcement Learning: An Industry Perspective
Building operations contribute approximately 28% of global greenhouse gas emissions according to the International Energy Agency. With the increase in cooling demand due to rising global temperatures, the optimization of rooftop units (RTUs) in buildings becomes crucial for reducing emissions. We focus on the optimization of the economizer logic within RTUs, which balances the mix of indoor and outdoor air. By effectively utilizing outside air, RTUs can significantly decrease mechanical energy usage, leading to reduced energy costs and emissions. However, the current practice of economizer optimization relies on static guidelines set by ASHRAE, which approximates the dynamics of individual facilities. We introduce a reinforcement learning (RL) approach that adaptively controls the economizer based on the unique characteristics of individual facilities. We have deployed our solution in the real-world across a distributed building stock. We address the scaling challenges with our cloud-based RL deployment on 10K+ RTUs across 200+ sites.
- Open Access
- 16 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 242-265
Rule-based Policy Regularization for Reinforcement Learning-based Building Control
Rule-based control (RBC) is widely adopted in buildings due to its stability and robustness. It resembles a behavior cloning methodology refined by human experts; however, it is incapable of adapting to distribution drifts. Reinforcement learning (RL) can adapt to changes but needs to learn from scratch in the online setting. On the other hand, the learning ability is limited in offline settings due to extrapolation errors caused by selecting out-of-distribution actions. In this paper, we explore how to incorporate RL with a rule-based control policy to combine their strengths to continuously learn a scalable and robust policy in both online and offline settings. We start with representative online and offline RL methods, TD3 and TD3+BC, respectively. Then, we develop a dynamically weighted actor loss function to selectively choose which policy for RL models to learn from at each training iteration. With extensive experiments across various weather conditions in both deterministic and stochastic scenarios, we demonstrate that our algorithm, rule-based incorporated control regularization (RUBICON), outperforms state-of-the-art methods in offline settings by and improves the baseline method by in online settings with respect to a reward consisting of thermal comfort and energy consumption in building-RL environments.
- 16 June 2023
- Published by Association for Computing Machinery (ACM)
- p. 242-265
Open AccessRule-based Policy Regularization for Reinforcement Learning-based Building Control
Rule-based control (RBC) is widely adopted in buildings due to its stability and robustness. It resembles a behavior cloning methodology refined by human experts; however, it is incapable of adapting to distribution drifts. Reinforcement learning (RL) can adapt to changes but needs to learn from scratch in the online setting. On the other hand, the learning ability is limited in offline settings due to extrapolation errors caused by selecting out-of-distribution actions. In this paper, we explore how to incorporate RL with a rule-based control policy to combine their strengths to continuously learn a scalable and robust policy in both online and offline settings. We start with representative online and offline RL methods, TD3 and TD3+BC, respectively. Then, we develop a dynamically weighted actor loss function to selectively choose which policy for RL models to learn from at each training iteration. With extensive experiments across various weather conditions in both deterministic and stochastic scenarios, we demonstrate that our algorithm, rule-based incorporated control regularization (RUBICON), outperforms state-of-the-art methods in offline settings by and improves the baseline method by in online settings with respect to a reward consisting of thermal comfort and energy consumption in building-RL environments.