The growing need for sustainable energy practices necessitates technology-driven interventions that can effectively bridge the disparity between consumer intentions and actual behavior. This paper formulates and empirically substantiates an IoT-enabled digital nudge architecture designed to promote sustainable energy behavior. The architecture provides goal-setting, social comparison, feedback, and informational nudges across multiple digital channels, utilizing linked devices, data processing layers, and a rule-based nudge engine. An 815-responder survey was analyzed using structural equation modeling with partial least squares (SEM-PLS) to identify the drivers of sustainable energy behavior and explore technology readiness as a moderating factor. The results show that nudges utilizing the Internet of Things (IoT) significantly enhance the alignment between intention and behavior. Goal-setting and feedback mechanisms have the highest effects. The findings also demonstrate that being ready for new technology improves nudge response, highlighting the importance of user-centered system design. This paper presents a scalable infrastructure for integrating IoT into sustainability projects, as well as theoretical contributions to technology adoption and behavioral intervention research. The study enhances the dialogue on environmental technology by illustrating the implementation of digital nudges through IoT infrastructures to expedite progress toward the Sustainable Development Goals (SDGs).

The rapid growth of digital commerce has created an even bigger appetite for operational intelligence that is sustainable, transparent, and carbon-efficient. This work proposes AGIA² (AI-driven Green Intelligence and Audit), an AIoT-infused sustainable commerce architecture that integrates adaptive green intelligence with autonomous audit pipelines to offer real-time environmental optimizations and Environmental, Social, and Governance (ESG)-compliant decisions. The method utilizes a mixed multi-modal dataset comprising e-commerce transaction logs, IoT supply chain sensor streams, carbon intensity time series, and audit/ESG compliance indicators. Such a dataset allows extensive modeling of the digital-physical-environmental interactions in modern commerce ecosystems. The proposed solution uses machine learning such as SVM, RF, and LR; deep learning like CNN and RNN/LSTM; and hybrid ensemble to derive sustainability scores, detect anomalies, quantify the operational emissions, and automate compliance. Adaptive Green Intelligence Module: This does carbon forecasting, energy optimization, and dynamic inference routing between edge-cloud layers. Autonomous Audit Pipeline will handle real-time audit checks, ESG rule validations, and explainable anomaly detection via SHAP/LIME. The proposed AGIA² framework demonstrates superior performance compared to baseline models, including standalone Support Vector Machine (SVM), Random Forest (RF), Logistic Regression (LR), and Convolutional Neural Network (CNN) models, across key evaluation metrics such as accuracy, precision, recall, and F1-score in experiments with an accuracy of 96.1%, an F1-score of 96.0%, and a ROC-AUC of 0.98. It greatly decreases the false negatives of sustainability violations and increases audit confidence ratings. In the correlation and confusion matrix studies, there are significant dependencies between IoT activity, carbon intensity, and sustainability categories. The AGIA² framework advances sustainable commerce towards a unified, real-time, AIoT-enabled method to run businesses sustainably, be carbon emission conscious, and audit themselves. Unlike fragmented AI-only or IoT-only approaches, AGIA² embeds sustainability scoring and ESG-aligned auditability as core system functions rather than post hoc reporting layers.

Supply chain information systems increasingly need to support decisions, yet Internet of Things (IoT) and blockchain approaches remain largely data-centric. We propose a decision-centric system in which IoT data acquisition, an artificial intelligence decision engine and a blockchain layer form one closed loop: each daily routing plan is optimized jointly for cost, lateness and delivered demand, validated against shared partner constraints and recorded before execution. We evaluated it by simulation on the supply chain logistics problem dataset (41 real customers) over 30 paired seeds. A learned deep Q-network engine with decision-time rollout reduced cost per 1000 units delivered by 24.3% relative to a demand-aware heuristic and 28.0% relative to OR-Tools, deciding in 43 ms; seeding adaptive large neighborhood search (ALNS) with its plan gave the lowest cost, 28.0% below the heuristic and below ALNS alone in 29 of 30 seeds. Joint routing–demand optimization lowered cost per unit by 33–45% relative to a cost-only objective. Under hard time windows, a heterogeneous fleet and eleven other scenarios, rollout stayed 12–32% cheaper than the heuristic in all 130 paired runs, and seeded ALNS was cheapest in every setting. Evidence-based validation blocked every invalid plan with up to four of five validators faulty.