Scholarlink

Cyber-physical systems (CPS) have emerged as the operational backbone of Industry 5.0, where machine intelligence, human supervision, and ecological responsibility are expected to converge. Yet current deployments often struggle with inflexible decision loops, degraded robustness under noisy conditions, and a lack of explicit integration of sustainability goals. This gap limits the capacity of next-generation factories and energy infrastructures to adapt swiftly while remaining environmentally accountable. To counteract these shortcomings, we offer a hybrid soft computing framework that incorporates sustainability as a first-class optimization objective. The framework design integrates three complemented layers: first, fuzzy inference for resolution of uncertainty; second, neural forecasting for forecasting stability; and third, evolutionary optimization for trade-off decisions of multi-objectives. The novelty lies in the design of these modules into a functionally cohesive decision pipeline that reduces latency, while adapting to human-in-the-loop feedback models. This approach contrasts with earlier architectures, which typically optimized for efficiency alone or operated in siloed computational modes. Evaluation in a smart factory test environment confirmed the efficacy of the framework. The prediction accuracy increased, with a mean RMSE of 0.092 vs. 0.143 for baseline ANN, while the decision latency has dropped to 42 ms, which was a 23% reduction in decision latency relative to the GA/PSO controllers. There was a 16.4% increase in the Sustainability Performance Index (SPI) over conventional CPS, and an 11.7% smaller carbon footprint. This suggests that integrating soft computing synergies with explicitly ecological metrics can produce resilient, efficient, and human-aligned CPS. Therefore, the proposed model, along with having provided a theoretical contribution toward frameworks for soft computing paradigms of Industry 5.0, also provides a practical translation into a developmental pathway to sustainable Industry 5.0.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Next-Generation Soft Computing Frameworks for Sustainable Cyber-Physical Systems in Industry 5.0

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Divyanshu Sinha,
  • Rahul Koshti,
  • Anusha Preetham,
  • S. Pooja Nayak,
  • Udit Mamodiya,
  • Parul Goyal

摘要

Cyber-physical systems (CPS) have emerged as the operational backbone of Industry 5.0, where machine intelligence, human supervision, and ecological responsibility are expected to converge. Yet current deployments often struggle with inflexible decision loops, degraded robustness under noisy conditions, and a lack of explicit integration of sustainability goals. This gap limits the capacity of next-generation factories and energy infrastructures to adapt swiftly while remaining environmentally accountable. To counteract these shortcomings, we offer a hybrid soft computing framework that incorporates sustainability as a first-class optimization objective. The framework design integrates three complemented layers: first, fuzzy inference for resolution of uncertainty; second, neural forecasting for forecasting stability; and third, evolutionary optimization for trade-off decisions of multi-objectives. The novelty lies in the design of these modules into a functionally cohesive decision pipeline that reduces latency, while adapting to human-in-the-loop feedback models. This approach contrasts with earlier architectures, which typically optimized for efficiency alone or operated in siloed computational modes. Evaluation in a smart factory test environment confirmed the efficacy of the framework. The prediction accuracy increased, with a mean RMSE of 0.092 vs. 0.143 for baseline ANN, while the decision latency has dropped to 42 ms, which was a 23% reduction in decision latency relative to the GA/PSO controllers. There was a 16.4% increase in the Sustainability Performance Index (SPI) over conventional CPS, and an 11.7% smaller carbon footprint. This suggests that integrating soft computing synergies with explicitly ecological metrics can produce resilient, efficient, and human-aligned CPS. Therefore, the proposed model, along with having provided a theoretical contribution toward frameworks for soft computing paradigms of Industry 5.0, also provides a practical translation into a developmental pathway to sustainable Industry 5.0.