Conversational Demand Response: Bidirectional Aggregator-Prosumer Coordination through Agentic AI

Reda El Makroum; Sebastian Zwickl-Bernhard; Lukas Kranzl; Hans Auer

cs.AIMar 6, 2026

Conversational Demand Response: Bidirectional Aggregator-Prosumer Coordination through Agentic AI

Authors: Reda El Makroum, Sebastian Zwickl-Bernhard, Lukas Kranzl, Hans Auer

Organizations: Energy Economics Group, TU Wien, Gußhausstraße 25–29, Vienna, Austria · Industrial Economics and Technology Management, NTNU, Trondheim, Norway

Abstract

Residential demand response depends on sustained prosumer participation, yet existing coordination is either fully automated, or limited to one-way dispatch signals and price alerts that offer little possibility for informed decision-making. This paper introduces Conversational Demand Response (CDR), a coordination mechanism where aggregators and prosumers interact through bidirectional natural language, enabled through agentic AI. A two-tier multi-agent architecture is developed in which an aggregator agent dispatches flexibility requests and a prosumer Home Energy Management System (HEMS) assesses deliverability and cost-benefit by calling an optimization-based tool. CDR also enables prosumer-initiated upstream communication, where changes in preferences can reach the aggregator directly. Proof-of-concept evaluation shows that interactions complete in under 12 seconds. The architecture illustrates how agentic AI can bridge the aggregator-prosumer coordination gap, providing the scalability of automated DR while preserving the transparency, explainability, and user agency necessary for sustained prosumer participation. All system components, including agent prompts, orchestration logic, and simulation interfaces, are released as open source to enable reproducibility and further development.

Figures & tables

Fig. 1: Bidirectional communication in CDR.

Capability	Description
Real-time Coordination	Aggregator adapts requests to prosumer context rather than issuing static commands
Portfolio Planning	Aggregator communicates upcoming events, enabling prosumers to plan ahead
Performance Feedback	Prosumers receive quantified impact of past participation, reinforcing engagement

TABLE I: Downstream Capabilities Enabled by CDR

Capability	Description
Transparent Participation	Prosumers see the full impact and compensation in plain language before committing
Explainable Compensation	Rewards are explained per load and linked to market conditions
Adaptive Preference Setting	Prosumers can update constraints and preferences at any time, not only at enrollment

TABLE II: Upstream Capabilities Enabled by CDR

Fig. 2: Illustration of the dual-solve procedure for a representative household day. (a) Cost-optimized self-consumption baseline. (b) Schedule with DR commitment.

Direction	Scenario	Iterations	Tool calls	Tokens (k)	Time (s)
Downstream	Acceptance	3.6 $\pm$ 0.5	2.6 $\pm$ 0.5	23.4 $\pm$ 4.0	8.3 $\pm$ 2.1
	Rejection	5.0 $\pm$ 0.7	3.6 $\pm$ 0.5	34.2 $\pm$ 6.0	9.8 $\pm$ 1.9
	High-target	3.4 $\pm$ 0.5	2.4 $\pm$ 0.5	21.8 $\pm$ 4.1	7.8 $\pm$ 2.2
Upstream	Availability	1	0	1.3 $\pm$ 0.0	1.3 $\pm$ 0.8
	Preference	1	0	1.0 $\pm$ 0.0	1.7 $\pm$ 1.6
	New Asset	1	0	1.6 $\pm$ 0.1	1.3 $\pm$ 1.2

TABLE III: CDR Computational Feasibility (mean

\pm

std,

n{=}5

)

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Parameter	Value
Battery
Capacity	15 kWh
Charge/discharge rate	8 kW
Round-trip efficiency	92%
SoC bounds	20–90%
Initial SoC	30% (4.5 kWh)

Appendix

TABLE IV: Simulation Parameters

Demand Response

Demand response shifts or reduces electricity use to help balance the grid while accounting for prices, incentives, comfort, and service needs.

See all 23 papers

LLM-Mediated Demand Response Coordination in Smart Microgrids
Jun 9, 2026J. de Curtò, I. de ZarzàMulti-Agent Coordination Agent-Based Modeling
Department of Computer Applications in Science & Engineering, BARCELONA Supercomputing Center, 08034 Barcelona, Spain · Escuela T´ecnica Superior de Ingenier´ıa (ICAI), Universidad Pontificia Comillas, 28015 Madrid, Spain · Human Centered AI, Data & Software, LUXEMBOURG Institute of Science and Technology, L-4362 Esch-sur-Alzette, Luxembourg
EnergyBridge: Benchmarking Household Energy Management, User Participation, and Grid Flexibility
Aug 9, 2026Xudong Wu, Zeqing Wu, Jiarui Zhang +8Human-in-the-Loop Evaluation Building Energy Management
1The University of Hong Kong · 2Tianjin University · 3Beihang University
Capacity-constrained demand response in smart grids using deep reinforcement learning
Feb 18, 2026Shafagh Abband Pashaki, Sepehr Maleki, Amir BadieeDemand Response Constrained RL
Lincoln AI Lab, University of Lincoln, Lincoln, United Kingdom

Jun 9, 2026J/K move · Enter open · S save

LLM-Mediated Demand Response Coordination in Smart Microgrids

J. de Curtò, I. de Zarzà

Department of Computer Applications in Science & Engineering, BARCELONA Supercomputing Center, 08034 Barcelona, Spain · Escuela T´ecnica Superior de Ingenier´ıa (ICAI), Universidad Pontificia Comillas, 28015 Madrid, Spain · Human Centered AI, Data & Software, LUXEMBOURG Institute of Science and Technology, L-4362 Esch-sur-Alzette, Luxembourg

Multi-Agent Coordination Agent-Based Modeling Demand Response LLM Agents Multi-Agent Communication Power Systems

Open paper AskPDF arXiv

Effective demand response in smart microgrids requires prosumers to cooperate voluntarily under strategic self-interest, a coordination problem structurally equivalent to a repeated Prisoner's Dilemma on a social network. This paper presents a multi-agent simulation in which a Large Language Model (LLM) Influence Compiler issues structured demand-response directives to a population of heterogeneous prosumer agents, each governed by a hybrid decision architecture combining game-theoretic base probability (derived from payoff history, neighbour imitation, and exploitation memory) with LLM narrative evaluation of incoming coordination signals. The hybrid architecture resolves a key methodological challenge: LLMs aligned via Reinforcement Learning from Human Feedback (RLHF) exhibit strong cooperation bias when used as direct decision-makers, producing flat dynamics regardless of grid conditions. By separating strategic reasoning from grounded narrative evaluation, the model generates realistic prosumer behaviour across six personality archetypes, with baseline cooperation near 50% and clear differentiation under influence. Compiled structured directives achieve 33.3% demand-curtailment cooperation versus 27.0% for unstructured messaging and 28.0% for a no-intervention baseline (

Δ_\mathrm{comp} = +0.063

), with the advantage preserved across both grounded and idealized agent substrates (

Δ= +0.083

) and across all resistance levels (

R = 0.1

0.7

). Hub-targeted dissemination via high-centrality network nodes outperforms peripheral or random targeting, confirming that grid topology provides mechanistic amplification independent of message content. These results suggest that structured LLM compilation, grounded agent reasoning, and network-aware targeting are complementary design principles for scalable, interpretable demand-response coordination in smart-city energy systems.

Abstract

Demand Response

Explore similar work

LLM-Mediated Demand Response Coordination in Smart Microgrids

EnergyBridge: Benchmarking Household Energy Management, User Participation, and Grid Flexibility

Capacity-constrained demand response in smart grids using deep reinforcement learning