In this second part of the special issue, we extend the conversation beyond the foundations established in Part I, which highlighted the transformative potential of generative AI and large language models for knowledge reuse, design exploration, data fusion, and smart manufacturing operations. Articles in this second part offer insights along three complementary themes. The first theme explores how generative AI structures and manages engineering knowledge and risk. This involves leveraging large language models (LLMs), vision language models (VLMs), and knowledge graphs to turn unstructured documentation, assembly procedures, and historical recall records into actionable, validated information assets. The second theme answers how generative and deep learning models reshape design and manufacturing systems. Articles in this theme investigate new generative and deep learning approaches for geometry synthesis, control, and process monitoring that operate over high-dimensional design and signal spaces. The third theme studies how human designers collaborate with AI in creative and decision-intensive contexts. Articles in this theme examine AI in collaborative roles such as standing in for interview participants, participating in speculative design, and narrating trade-offs in multi-objective manufacturing decisions. This group of articles simultaneously introduces interesting questions regarding bias, trust, and ethics.
Wei Chen, V. Krishnamurthy, Yanglong Lu et al.· Journal of Computing and Inf...· 0 citations
Visual fluency in generated video does not imply physical reliability, and a scalar quality score alone is incapable of indicating the obligation a clip violates or the moment it fails. We present VeriPhy, an auditable physical-verification system in which a text-only planner compiles the prompt into typed physical obligations and a statically validated execution plan before any frame is observed. During execution, observations gate and scope only declared calls to frozen low-level experts (e.g., segmentation and tracking, counting, eleven typed physical measurements over the resulting tracks, depth, OCR, and audio-event detection). Each action returns a provenance-carrying evidence record whose payload, when usable, is either a typed measurement or an explicitly tagged learned state. Typed resolvers and fixed composition map usable records to a three-valued state (supported, contradicted, or unknown, surfaced as plausible, implausible, or abstain) with full provenance, so that every verdict is traceable to the evidence that produced it. We anchor evaluation in a 1,500-clip corpus of human-annotated flaw records that localize real generation failures in prompt reference, space, and time. On a 149-clip core carrying 304 such records, VeriPhy accounts for 228, against 164 for a published question-decomposition evaluator given the same clips and the same claims. Recall alone does not separate it from prompting the same backbone monolithically, which reaches 222; what separates them is that each decision retains its evidence record and provenance, making the traces auditable one verdict at a time and usable as the interface through which a critic verdict could be written back into generation.
Wenzhuo Xu, Yu-Chen Zhu, Chongjian Ge et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
