Siemens and Microsoft Leverage Generative AI to Elevate Industrial Operations

Technology and software giants Siemens and Microsoft have announced their latest partnership to enhance industrial operations with generative artificial intelligence (AI).

The collaboration will combine a product lifecycle management (PLM) software from Siemens (called Teamcenter) with Microsoft’s collaborative business platform, Microsoft Teams, and the natural language models within the Azure OpenAI Service.

 

Customers can use the Teamcenter app on mobile devices to access Azure OpenAI Service. Image used courtesy of Pixabay

 

Enhancing Collaboration and Reporting

Using the new Teamcenter app for Microsoft Teams, frontline workers, design engineers, and collaborative teams within specific business fractions can access Azure OpenAI Service through mobile devices.

Users can record and report quality or product design issues with Azure OpenAI Service natural language models (such as GPT-2, GPT-3, and BERT). The models can compile and translate speech data into insightful reports, which can then be passed on through the Teamcenter app to the most suitable manufacturing, design, or engineering professional.

Concerning speech input, different language options are available to workers to make the Teamcenter app accessible to all. The Teamcenter app will become available to users later this year, providing access to PLM tools to enhance engineering designs and manufacturing processes.

 

Visual Inspection and Quality Control

Another way generative AI can engender advanced digital transformation of factory automation operations is through visual inspection and quality control.

Generative AI can scan physical products and/or images to detect any faults or issues that human operators might miss. AI models learn from exposure to big data sets to pick up on certain characteristics or patterns that suggest a product or machine system has a defect or anomaly.

Similarly, generative AI models can scan products during different manufacturing stages to ensure quality standards are being met. Preset criteria or a reference image are typically used to assess and confirm quality control measures.

 

ChatGPT can use historical PLC data to predict machine behavior. Image used courtesy of Pixabay

 

Another key aspect of factory automation operations is predictive maintenance. AI models can help analyze product/machinery images to assess asset health: determining the level of damage or wear that could lead to serious faults and unwanted downtime. If asset health is less than optimum, the AI can immediately alert maintenance staff to fix the problem before it is too late. Predictive maintenance is key to keeping overall production costs down.

By combining the capabilities of Siemens’ Industrial Edge and Microsoft’s Azure Machine Learning (ML), ML systems can analyze shop floor videos and/or images to create, run, and track AI vision models.

 

Software Engineering

At the international industry development-focused trade show, Hannover Messe, Microsoft, and Siemens showcased what is possible when using ChatGPT and other AI services from Azure to enhance Siemens’ engineering solutions for industrial automation applications. Automation engineers and software developers can speed up the process of generating PLC code by using a natural language input as a baseline.

For example, developers can use ChatGPT to create code snippets (or short stretches of reusable text, source, or machine code) for larger programming tasks or recurrent PLC programming exercises. ChatGPT can also imbibe PLC historical data for training and use it to predict aspects of machine behavior.

Azure Cognitive Services and Azure ML can also help analyze PLC data to assess performance and allow for further optimization through continued PLC code development.