ChatGPT in the current manufacturing landscape

How ChatGPT Helps Manufacturers

An early use case is to help manufacturers perform text-based tasks.

Bret Greenstein, cloud and digital data and artificial intelligence partner at professional services firm PwC, said: “Most product manufacturing processes are handled by massive text-based workflows and documents. , from machine maintenance logs and service entries to status reports and alerts.

Additionally, manufacturers use document-heavy workflows for parts orders, material movement, shift assignments for workers, and logs. All these business elements can be analyzed and generated with the help of generative AI.

ChatGPT's capabilities go beyond rote tasks. Generative AI applies across the entire manufacturing value chain: market research, product conception, design, engineering, and supply chain management. It could also benefit from better product configuration and recommendation engines.

According to Raghuram Mocherla, vice president of Capgemini Engineering, a global engineering consultancy, the top manufacturing use cases for ChatGPT include the following:

• Integrate unstructured market demand, changes in the regulatory environment, supply change constraints, and other inputs into product design.
• Document and apply problem solving methods used by engineers and designers during the product design and manufacturing phases.
• Construction engineering natural language amplifies B2B engine configuration, price and offer functions.
• Automate product testing and validation by generating test scenarios and analyzing the results.
• Assist in purchasing decisions based on cost, supply chain constraints, and production capacity.
• Link product performance and field issues to flaws in product design and storage.

Risks and Challenges of Using ChatGPT in Manufacturing

Manufacturers face many risks and challenges when making large language models (LLMs) loose in the factory.

Most worrying are hallucinations, says Beena Ammanath, executive director of the Deloitte AI Institute. Inaccurate and imagined information has no place in a manufacturing company or other business.

Another problem is that generative AI models are limited by their training data. Integrating additional datasets requires costly retraining and fine-tuning. Ammanath said these limitations could jeopardize core applications such as data management and predictive maintenance. The consequences can be as severe as machine downtime and failure.

Also consider the legal risks associated with ChatGPT in manufacturing. Companies must ensure that engineers do not inadvertently expose confidential information to these services. Samsung bans use of ChatGPT 2023 after discovery of data breach. Manufacturers could face potential liability for unauthorized use of models based on data input. Assigning liability may create similar legal risks when an AI model directly or indirectly causes bodily harm to equipment or personnel. Therefore, when ChatGPT and similar applications are adopted in manufacturing, the security, information security and privacy risks associated with processing and analyzing human data must be explored.

Best Practices for Adopting ChatGPT in Manufacturing

Kamalesh?, Global Head of AI Practice, Business Transformation Group, IT services and consulting firm Tata Consultancy Services, recommends careful use case and data planning, security design, continuous monitoring, transparency mechanisms, user training and risk management planning.

Since generative AI tools rely on data for training, it is also crucial to focus on improving the data. "Success will still come down to streamlining processes, data, and systems of record to provide reliable, diverse, and statistically significant training datasets," Mocherla said. Additionally, the manufacturing and business-specific language used as ChatGPT prompts must be standardized.

Technologists in manufacturing must work with AI experts. “For generative AI tools to work well, organizations need people who understand the technical details of those tools and work closely with people who understand the business processes that AI is enhancing,” Greenstein said.

Much of this work is accessing the right data or examples of past work to train the LLM and drive the generative AI to produce accurate results. Obtain input from business parties and manufacturing workers during the process to assess the quality and accuracy of the data or responses.

Manufacturing will inevitably adapt to new trends

As we've seen with recent iterations of GPT, generative AI models will get even bigger. They will include more parameters, as well as multimodal and multilingual capabilities. They are also working on more integration with other technologies, such as computer vision, IoT and robotics.

Greenstein predicts that new innovations will shift the technology from merely responding to prompts to using prompts and responses like an agent to achieve business goals. These developments will expand the types of work that generative AI can do, and expand the degree of productivity and impact it can have. If these predictions play out, ChatGPT capabilities could revolutionize the dusty old UIs currently found in manufacturing equipment, Mocherla said. A more natural human-machine interface will capture unstructured information so it can be incorporated into the process without a lot of effort.