AVAlex Vlachokostas

Alex Vlachokostas

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

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Alex Vlachokostas has a total of 57 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Addressing the split incentive challenge for rooftop solar PV and battery energy storage in multifamily rental buildings (CRADA 638) Final Report

    by Efthimios Alexandros Vlachokostas, Usdoe, Alexander C. Buchko, Roshan L. Kini, Milan Jain, Eliza Tobin, Charlotte Meerstadt

    This project advances the understanding of how roof solar PV systems and battery energy storage systems (BESS) can be effectively deployed in multifamily residential buildings, a sector that has historically faced barriers due to misaligned incentives between landlords and tenants. By leveraging high-resolution building stock data and simulation tools, the research demonstrates how energy consumption patterns vary across building types, climates, and occupant characteristics, and how these variations influence the optimal sizing and operation of distributed energy resources. A key contribution is the development of a publicly accessible, web-based tool named RESIDE (Residential Energy Systems & Infrastructure Data Evaluation) that allows users to explore building energy use and evaluate solar and battery configurations without requiring specialized expertise. This significantly lowers the barrier to entry for stakeholders such as property owners, utilities, and policymakers. From a technical perspective, the project shows that integrating rooftop solar PV with battery energy storage can substantially reduce electricity costs and peak demand through strategies such as energy arbitrage and peak shaving. The modeling framework incorporates real-world constraints, including time-of-use electricity pricing and battery degradation, providing realistic and actionable insights. Economically, the results indicate that properly sized systems can deliver meaningful cost savings, improving the feasibility of energy investments in multifamily housing. More broadly, the project benefits the public by supporting the transition to affordable and reliable energy, particularly in rental multifamily housing where adoption has traditionally lagged.

  • Open Access

    Experimental evaluation of multi-agent reinforcement learning for corridor-level traffic signal control

    by Roshan L. Kini, Alex Vlachokostas, Salman Shuvo, Sayak Mukherjee, Samrat Chaterjee, Nawaf Nazir, Sonja Glavaski, Andrew Valdez, Sakshat Katyarmal, Ning Yang, et al.

    Although recent research has advanced adaptive traffic control, significant challenges remain in optimizing signal phase timing, coordinating across heterogeneous multi-intersection corridors, and validating performance under real-world traffic conditions. This paper presents a signalized multi-agent reinforcement learning (MARL) approach and validates its performance within a novel hardware-in-the-loop (HiL) framework, using actuated control as the baseline for comparison. The proposed framework integrates a real-world traffic controller with a digital twin calibrated simulation model, enabling real-time communication between simulation and traffic control infrastructure. The MARL algorithm runs on a roadside unit edge device and updates signal phases across three intersections. Both single and dual-ring signal configurations are evaluated within the framework to reflect typical real-world traffic configurations. The digital twin is calibrated using traffic data from the City of Coral Gables, Florida, USA. The MARL-informed HiL results show a 31% reduction in travel delay (sec/mi) in the morning and a 46% reduction in the afternoon, indicating significant improvements in traffic flow efficiency during peak periods compared to actuated control.

  • Open Access

    Experimental evaluation of multi-agent reinforcement learning for corridor-level traffic signal control

    by Roshan L. Kini, Alex Vlachokostas, Salman Shuvo, Sayak Mukherjee, Samrat Chaterjee, Nawaf Nazir, Sonja Glavaski

    Although recent research has advanced adaptive traffic control, significant challenges remain in optimizing signal phase timing, coordinating across heterogeneous multi-intersection corridors, and validating performance under real-world traffic conditions. This paper presents a signalized multi-agent reinforcement learning (MARL) approach and validates its performance within a novel hardware-in-the-loop (HiL) framework, using actuated control as the baseline for comparison. The proposed framework integrates a real-world traffic controller with a digital twin calibrated simulation model, enabling real-time communication between simulation and traffic control infrastructure. The MARL algorithm runs on a roadside unit edge device and updates signal phases across three intersections. Both single and dual-ring signal configurations are evaluated within the framework to reflect typical real-worldtraffic configurations. The digital twin is calibrated using traffic data from the City of Coral Gables, Florida, USA. The MARL-informed HiL results show a 31% reduction in travel delay (sec/mi) in the morning and a 46% reduction in the afternoon, indicating significant improvements in traffic flow efficiency during peak periods compared to actuated control.

  • Open Access

    Experimental Demonstration of Transactive Coordinated Control for Residential Demand Response

    by Roshan L. Kini, Alex Vlachokostas, Michael R. Brambley, Robert Lutes, Andrew Costinett, Ke Ma, Austin Rogers

    Load flexibility is essential for balancing supply and demand during peak price periods, supporting grid stability and lowering energy costs. However, traditional utility-led direct load control programs have been criticized for occupant discomfort and declining participation. In contrast, the proposed transactive method accounts for household preferences while intelligently balancing comfort with cost and coordinating the operation of in-home appliances. This paper presents the methodology and test results for a transactive coordinated load control of space cooling and water heating to respond to time-of-use and real-time electricity pricing. The method employs data-driven models for both cooling and water heating to show that effective demand response can be achieved at short control intervals (10~min), with demand curves generated dynamically at each interval and aggregated to determine a clearing price that coordinates power allocation while maintaining occupant comfort. The algorithm was implemented and validated in a physical home using commercially available internet-connected thermostats and water heater controllers, demonstrating that for time-of-use pricing, the strategy achieved 41.7\% demand reduction; for real-time pricing, 32.5\% demand reduction while maintaining temperatures within the comfort range.

  • Open Access

    Addressing the Split Incentive Challenge for Enhanced Solar Adoption in Multifamily Rental Properties (CRADA #638) Abstract

    by Usdoe, Efthimios Alexandros (Alex) Vlachokostas, Roshan L. Kini, Milan Jain, Alexander C. Buchko, Jennifer L. Fanning, Rachel B. Thompson

    The split incentive problem is particularly pronounced in rental markets, where landlords prioritize investments that directly increase property value or rental income. Since energy savings from solar photovoltaic (PV) systems primarily benefit tenants, landlords may perceive little return on investment unless mechanisms exist to recapture some of the financial gains. The primary objective of this project is to develop a publicly available, web-based tool to analyze the U.S. Department of Energy’s ResStock database, which models the U.S. residential building stock. The tool allows users to filter buildings by location, type, HVAC system, square footage, and other characteristics, and outputs typical electric load profiles. By leveraging location-specific electric load data, Fram Energy aims to advance business strategies that address the split incentive barrier and promote the adoption of solar PV installations in rental properties. In addition, a machine learning model will be developed to weigh the marginal contribution of building features across the dataset in predicting electricity demand, supporting guided decision making in forecasting electric load profiles. Lastly, based on each building’s location, load profile, and utility’s electricity rate, an optimized solar photovoltaic array and battery energy storage system will be sized to provide energy arbitrage opportunities.

  • Open Access

    Occupant-Centric Demand Response for Thermostatically-Controlled Home Loads

    by Roshan L. Kini, Alex Vlachokostas, Michael R. Brambley, Austin Rogers

    Efficiently managing energy usage to balance supply and demand on the electric grid is crucial, especially with the widespread deployment of distributed variable renewable electricity generation. This paper introduces two duty-cycle control methods for heating systems, adjusting thermostat setpoints to limit and shift electricity demand. The control approaches employ innovative techniques, such as adaptive duty cycling, to prioritize household thermal comfort while reducing peak demand. These control methods can respond to signals from the electric grid, including demand targets and time-of-use tariffs, and were tested physically on an electric furnace and heat pump in a test home during winter conditions in 2021 and 2022. The results are given as average demand reductions and energy use impacts with respect to the average indoor-outdoor temperature difference during the control period. For heat pumps, demand limiting control reduced power by 18.5% and 23.3% for indoor-outdoor temperature differences of 30.F and 40.F. Preheating-based demand shifting achieved reductions of 34.8% and 33.2% for the same temperature differences. Electric furnace tests showed demand reductions of 33.8% and 25.3% for demand limiting, and 56.1% and 45.7% for preheating-based demand shifting. These findings highlight the potential for innovative control methods to enhance grid efficiency and reduce energy consumption.


  • Open Access

    National Cost-Effectiveness of ANSI/ASHRAE/IES Standard 90.1-2022

    by Usdoe, Matthew T. Tyler, Efthimios Alexandros Vlachokostas, Douglas E. Maddox, YuLong Xie, Michael I. Rosenberg, Michael Myer

    The purpose of this analysis is to examine the cost-effectiveness of the 2022 edition of ANSI/ASHRAE/IES Standard 90.1, which is developed by the ASHRAE Standard Standing Project Committee (SSPC) 90.1 and is the model energy standard for all commercial buildings and multifamily residential buildings over three floors. PNNL analyzed the cost-effectiveness of changes in Standard 90.1-2022, compared to the previous 90.1-2019 edition, as applied in commercial buildings across the United States. In reviewing proposed changes to Standard 90.1, the SSPC considers the cost-effectiveness of individual changes (addenda). Due to the continuous nature of the development process, however, ASHRAE does not evaluate the entire package of addenda from one edition of the standard to the next, which is of particular interest to adopting state and local governments. Providing states with an analysis of cost-effectiveness facilitates a more comprehensive understanding of the impacts associated with updated model energy codes, informs the state decision-making process and its authorities, and ultimately encourages greater adoption of updated energy codes. This information also informs the development of future editions of Standard 90.1.