{"id":2914,"date":"2020-01-22T07:21:17","date_gmt":"2020-01-22T12:21:17","guid":{"rendered":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/?p=2914"},"modified":"2026-03-26T12:02:20","modified_gmt":"2026-03-26T16:02:20","slug":"the-possibilities-and-limitations-of-ai-and-the-digital-twin","status":"publish","type":"post","link":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/the-possibilities-and-limitations-of-ai-and-the-digital-twin\/","title":{"rendered":"Today meets tomorrow: The opportunities and limitations facing AI and the digital twin (Part II)"},"content":{"rendered":"\n

In our previous\nblog<\/a>, we discussed the challenges of simulation and the digital twin as\nwell as how artificial intelligence (AI) can bring real world data into the\ndigital twin. <\/p>\n\n\n\n

In this part, we will focus on\nthe use of AI on the factory floor and the limitations engineers could face\nwhen using AI to implement a digital twin.<\/p>\n\n\n\n

The potential uses for\nartificial intelligence are enormous, yet unexploited. While artificial\nintelligence can help build simulation models, many are using the technology to\ncollect and analyze vast amounts of data. Financial institutions use its\nalgorithms for fraud detection and manufacturers can review products for\ndefections and use consistently-flowing data to build a more robust digital\ntwin.<\/p>\n\n\n\n

AI and the digital twin on the factory floor<\/strong><\/h3>\n\n\n\n

Artificial intelligence<\/a> has tremendous potential not\nonly in designing products, but also when it comes to making production more\nefficient, flexible and reliable. As industry continues toward increased\ndigitalization, data in production environments are the basis on which entire\nplants operate and systems generate.<\/p>\n\n\n\n

AI can bring\ngreater reliability and efficiency to the factory floor<\/a>. For instance, unplanned\ndowntime has always been a major problem that industrial manufacturers have\nfaced. A digital twin can use AI-based analyses for predictive maintenance,\nwhich can reduce downtime by scheduling or alerting of maintenance needs before\nthey become problematic.<\/p>\n\n\n\n

Digital twins\nhave been used to structure the planning and design of products and machinery,\neven production operations themselves. This flexible and more efficient\napproach have helped manufacturers produce more affordable, high-quality and\ncustomized products faster.<\/p>\n\n\n\n

AI can now use\nthese machines and processes to gather insights from these high volumes of data\nby themselves and optimize their processes during live operation. On top of\nthat, neural processing units<\/a> are now part of controllers\nfor object recognition, visual\/sensor-supported quality checks in production plants and image-guided\nrobot systems are more flexible when reacting to unexpected situations. If a\nmachine-learning algorithm spots quality defects, it can respond automatically\nduring runtime.<\/p>\n\n\n\n

Edge computing can process data where it\u2019s generated, whether that\u2019s at the plant or the machine. One such location is the Siemens Electronics Works Amberg<\/a>, Germany, which produces Programmable Logic Controllers (PLCs) and other industrial automation products for the world market. To ensure quality, the printed circuit boards (PCBs), a critical part of the controllers, are tested through x-ray during the manufacturing process. The problem with X-ray testing is that it requires long throughput time and hence is the factory\u2019s bottleneck. An additional x-ray machine would result in capital investment of \u20ac500,000 and must be integrated into the manufacturing environment. Instead, AI experts trained an algorithm on an edge application at the plant to predict whether the soldered joints on the PLCs were free of faults; in other words, whether or not an end-of-line test is necessary. Thereby, test efforts for the x-ray machines were reduced by up to 30 percent in the first step. Closed-loop analytics are planned to be factored into production in the near future.<\/p>\n\n\n\n

\"\"
Amberg’s X-Ray<\/figcaption><\/figure>\n\n\n\n

Limitations of AI in the digital twin<\/strong><\/h3>\n\n\n\n

Regardless of its benefits, AI\ndoesn\u2019t come without its challenges. There are several shortcomings and\nlimitations that engineers are likely to encounter when using AI techniques to\nimplement a digital twin:<\/p>\n\n\n\n

  • AI\ntechnology needs trusted, \u2018clean\u2019 mass data for teaching the AI system with\nrespect to the expected properties\/behavior. This is often hard to get and to\nverify, especially for newly designed systems. On the other hand, the\nverification and certification of AI-based behavior in safety-critical\napplications, such as autonomous driving, is a challenge yet to be solved.<\/li>
  • A\ndigital twin mirrors the properties\/behavior of a concrete asset, while AI\nsystems are essentially mirroring statistical properties\/behavior. This can make\nAI twins problematic.<\/li>
  • AI\ntechnologies have no physical background. They can show deviations between real\nand expected behavior but cannot easily explain the physical reasons behind it.<\/li>
  • Finally, customers are always concerned with how\nthey can protect IP and the data associated with those assets while data\nsharing.<\/li><\/ul>\n\n\n\n

    There are two basic approaches to addressing the limitations when\ncombining conventional AI techniques with physics models:<\/p>\n\n\n\n

    • Sequential approaches<\/strong>. Using AI to detect anomalies before a simulation is employed for detailed analysis. Conversely, simulations are used to teach AI before it is applied as fast surrogate model for simulations.<\/li>
    • Parallel approaches<\/strong>. AI and simulation models are combined. For example, by substituting well understood components by AI surrogates in a system simulation, using AI to calibrate simulation models, or using AI to automatically analyze complex simulation results.<\/li><\/ul>\n\n\n\n

      Summary<\/strong><\/h3>\n\n\n\n

      There is\nmassive potential for growth, efficiency and innovation when using artificial\nintelligence. The potential applications for AI are monumental across\nindustrial manufacturing. This is merely the beginning of the opportunities\nthis technology will provide.<\/p>\n","protected":false},"excerpt":{"rendered":"

      In our previous blog, we discussed the challenges of simulation and the digital twin as well as how artificial intelligence…<\/p>\n","protected":false},"author":69388,"featured_media":2929,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"spanish_translation":"","french_translation":"","german_translation":"","italian_translation":"","polish_translation":"","japanese_translation":"","chinese_translation":"","footnotes":""},"categories":[1],"tags":[12,11,35],"industry":[],"product":[],"coauthors":[],"class_list":["post-2914","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","tag-artificial-intelligence","tag-digital-twin","tag-industrial-automation"],"featured_image_url":"https:\/\/blogs.sw.siemens.com\/wp-content\/uploads\/sites\/19\/2020\/01\/kv-futureofautomation-iso2_small.jpeg","_links":{"self":[{"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/posts\/2914","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/users\/69388"}],"replies":[{"embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/comments?post=2914"}],"version-history":[{"count":4,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/posts\/2914\/revisions"}],"predecessor-version":[{"id":2936,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/posts\/2914\/revisions\/2936"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/media\/2929"}],"wp:attachment":[{"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/media?parent=2914"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/categories?post=2914"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/tags?post=2914"},{"taxonomy":"industry","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/industry?post=2914"},{"taxonomy":"product","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/product?post=2914"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/blogs.sw.siemens.com\/thought-leadership\/wp-json\/wp\/v2\/coauthors?post=2914"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}