Eco-design integrates environmental considerations into product design, recognizing 80% of sustainability impacts determined at the design phase. Artificial intelligence (AI) provides powerful tools for optimizing designs, assessing environmental impacts, and supporting circular economy, making eco-design proactive. Despite AI use in sustainable product development, no review has synthesized these efforts. Therefore, we conducted a systematic review using the PRISMA method, covering 38 studies from 2014 to 2024 applied AI in eco-design. This is the first review to consider all life-cycle stages with eco-design practices, integrating Ellen MacArthur circularity principles, United Nations sustainable development goals (SDGs), life cycle assessment (LCA), industrial applications, and AI methods. Our findings reveal: 1- an imbalanced focus across product life-cycle stages, with most studies addressing design and end-of-life, while production, use-life, and distribution remain underexplored. 2- Common eco-design practices include recycling, energy reduction, and disassembly, with less focus on non-hazardous materials, waste minimization, and remanufacturing. 3- While neural networks and hybrid AI methods are commonly applied for material compatibility and emissions prediction, more advanced AI-based approaches such as generative AI and LLMs have yet to be used in design, LCA, and circularity analysis. 4- No study applies all four Ellen MacArthur Technosphere circular economy strategies. 5- Researchers rarely couple LCA with cradle-to-cradle assessments or embed their results in real-time design simulations. 6- Case studies mostly focus on electronics and household appliances, with limited application in automotive, aviation, maritime, and healthcare. 7- SDG consideration mainly centers on SDGs 12 and 13, with more attention needed for other SDGs.

Dynamic Binary Instrumentation (DBI) in user space often suffers from low probe insertion success rates and high execution overhead, due to challenges in handling the compact instruction layouts ($\lt $5 bytes) and complex trampoline placement constraints. Existing techniques are either limited in scope, incur high runtime overhead, or rely on heavyweight code relocation. This paper introduces NOProbe, a lightweight, user-space DBI framework that enables safe and efficient probe insertion using two novel strategies. The first strategy locates trampoline sites by leveraging compiler-generated NOP paddings; the second employs pseudo-NOP instructions to support trampoline placement even when instructions overlap. Additionally, we propose a thread-safe patching algorithm,lock-redirect-load-arm, for safe runtime code modification. Experimental results show that NOProbe achieves 97%-99% probe effectiveness, reduces probe insertion latency, and maintains very low per-probe execution overhead, even under high probe density and multithreaded workloads.

This paper presents a hybrid chatbot for sustainable cryptocurrency investment, powered by small-scale Large Language Models (LLMs). We evaluate the performance of various lightweight LLMs (under 10B parameters) on blockchain and sustainability-related tasks, demonstrating that carefully orchestrated smaller models can effectively match or exceed the performance of larger models in domain-specific applications. Our multi-agent Retrieval-Augmented Generation (RAG) pipeline, incorporating real-time sustainability metrics from the Crypto Carbon Ratings Institute, achieved 87.09% accuracy in providing investment guidance, improving upon the 72.58% baseline of individual models. The results show that refined instruction engineering and specialized pipeline architecture can significantly enhance model performance without requiring larger, more energy-intensive models. This work contributes to both the practical implementation of sustainable cryptocurrency investment tools and the broader discussion of environmental considerations in AI system design and deployment.

Real-time freight matching and dispatching present complex challenges due to the dynamic nature of supply and demand, high-dimensional decision spaces, and the need for rapid response under operational constraints. In this paper, we propose a novel Hierarchical Reinforcement Learning (HRL) framework that jointly optimizes matching and dispatching processes in freight transportation, offering enhanced responsiveness, modularity, and coordination under real-time constraints. To mitigate computational demands without sacrificing matching accuracy, we present two efficient pre-filtering algorithms, PAMA (Pre-filtering Algorithm for Matching Agent) and PADA (Pre-filtering Algorithm for Dispatching Agent), which enhance the H3 hexagonal geospatial partitioning system. A GIS-based simulation of freight flow in Montreal provides realistic validation. Experimental results show that our framework improves successful match rates by 1.73%, reduces vehicle idle time by 5.7%, and maintains low empty mileage. Compared to state-of-the-art baselines including Deep Q-Network and clustering-based methods, our HRL model achieves superior reward efficiency, scalability, and adaptability in dynamic freight environments. These findings underscore the potential of HRL to enhance the operational efficiency of smart freight platforms.

This research presents a novel Deep Q-Learning (DQL) framework designed for efficient real-time matching of shipments and vehicles in the freight transportation sector. The framework utilizes the H3 geospatial indexing system for accurate positioning and employs a pre-filtering mechanism to streamline the matching process. When evaluated on a simulated model of Montreal’s transportation network, the framework demonstrates promising results in generating matches that reduce travel distance and prioritize timely service. Through extensive experimentation, a configuration utilizing ReLU activation was identified as particularly efficient, even under limited computational resources. This research contributes to the development of advanced, real-time matching algorithms in logistics and show-cases the potential of integrating reinforcement learning with geospatial analysis to address complex transportation challenges. These findings offer valuable insights for freight companies seeking to improve their matching processes, potentially leading to cost reductions and enhanced service quality.

The effective pedagogy strategy for teaching circular economy is essential. Industry 4.0 influenced educations in different ways. This working paper presents the perspectives of the application of digital twin and gamification in teaching circular economy. A preliminary conceptual framework is proposed for teaching End of Life (EoL) products recovery topic.