在工业领域，将人工智能的自主能力与有效的人类监督相结合正成为关键议题。据2026年制造领导委员会的一项调查显示，66%的制造商目前正在使用或计划部署代理型AI（agentic AI），同时有40%的企业预计将在2030年前实现某种程度的工厂自主化1。然而，这种对自动化潜力的追求伴随着显著的顾虑，德勤2025年的智能制造调查指出，65%的高管将包括业务中断在内的运营风险视为首要担忧1。在此背景下，AVEVA首席产品官Rob McGreevy倡导利用“活数字孪生”技术来提升工业自主性，并辅助人类进行控制1。
尽管存在风险考量，已有部分企业通过AI应用取得了实质性成果。雀巢在其Nesquik工厂利用AI优化干燥器流程，在一项标志性测试中成功降低了16%的能源使用，同时提升了产品质量和产量1。泰国SCG Chemicals则借助覆盖工程、运营和维护全周期的AI支持数字孪生系统，实现了99%的工厂可靠性，并在短短六个月内获得了九倍的投资回报1。这些案例表明，AI在实时监测、预测维护及闭环优化中的应用已初见成效1。
目前，行业对于AI改变生态系统合作伙伴间决策与问责制的看法仍较为保守。AVEVA与IMD联合开展的研究显示，43%的组织表示AI仅轻微地或未改变现有的决策机制，只有23%的组织报告了显著变化1。此外，监管环境也在塑造这一格局，欧盟《人工智能法案》明确将关键基础设施中的某些AI安全组件归类为高风险，并强制要求实施人工监督1。因此，尽管制造商对AI自主运行表现出浓厚兴趣，但出于风险控制的需要，界定操作边界并保留人类监督仍是当前的主流做法1。
Industrial manufacturers are increasingly integrating agentic artificial intelligence with "living digital twin" technologies to enhance operational autonomy, yet the industry remains cautious about fully removing human oversight. A 2026 survey by the Manufacturing Leadership Council revealed that 66% of manufacturers are currently using or planning to deploy agentic AI, while 40% anticipate achieving some level of factory autonomy by 2030 1. Despite this momentum, risk aversion persists; a 2025 Deloitte smart manufacturing study found that 65% of executives cite operational risks, including business interruptions, as their primary concern 1. Consequently, companies like AVEVA emphasize pairing AI capabilities with defined operational boundaries and human supervision rather than pursuing unrestricted autonomy 1.
Real-world applications demonstrate how AI can optimize specific processes without eliminating the need for human control in critical infrastructure. At Nestlé’s Nesquik plant, AI algorithms were used to optimize dryers, resulting in a 16% reduction in energy consumption alongside improvements in product quality and output during a landmark test 1. Similarly, SCG Chemicals in Thailand employed an AI-supported digital twin covering engineering, operations, and maintenance, which helped achieve 99% plant reliability and a ninefold return on investment within six months 1. However, the broader organizational impact of these technologies varies; research from AVEVA and IMD indicates that only 23% of organizations reported significant changes in decision-making and accountability among ecosystem partners due to AI, while 43% noted minor or no change 1. Regulatory frameworks also influence deployment strategies, with the European Union’s AI Act classifying certain safety components in critical infrastructure as high-risk and mandating human oversight 1.
评论
还没有评论，欢迎留下第一条。