Part 5: How to go beyond GenAI experimentation to deliver real value

In this latest article in our ongoing "Road to Industry 4.0" series, we will discuss the capabilities of GenAI and how it can be a game-changer across businesses in the manufacturing industry.

2023 saw the mass exploration of generative artificial intelligence (GenAI) across personal life and business. But if last year was mostly about experimentation across different sectors, 2024 is ushering in strategic use cases that can deliver real value, with companies shifting focus towards targeted investments.

Gen AI is more than a tech fad – it is the evolution of over 50 years of AI research, development and use cases. Underpinned by an increase in computing power, data integration and platform capability, the current wave has transformative potential – although it is not without risks.

Defining Generative AI

Generative AI refers to a category of artificial intelligence technology capable of creating diverse types of content, such as text, images, audio, and synthetic data. This is achieved through learning patterns from existing datasets, which GenAI then applies to produce new, original content.

A key factor fueling interest in the past year has been the development of straightforward and intuitive user interfaces that facilitate rapid content creation. This trend is particularly evident in platforms like OpenAI's ChatGPT, Google's Gemini, and similar applications, which have made content generation accessible and efficient for users.

GenAI in the context of manufacturing

The application of AI innovations in manufacturing (e.g. EdgeAI and Computer Vision) has been growing at a rapid pace, with significant advances in technology and adoption in recent years. The availability of vast amounts of data from Internet-of-Things (IoT) sensors, cheaper computing power via the cloud, as well as advanced deep learning algorithms have led to applications of these AI innovations in a number of use cases in the industry, such as predictive maintenance, quality control and process optimization, among others.

Alongside these other innovations, GenAI has now emerged as part of the AI portfolio for manufacturing. It offers the potential for improvement via further optimization and automation of work processes where there is need for text and image generation content.

Here are some examples of areas where GenAI can be deployed to give manufacturing organization a boost:

  • AI Co-Pilots – Training, Supply Chain Planning & Maintenance

Real-world examples: Microsoft, C3

In the area of supply chain management and plant maintenance, GenAI-based co-pilots are now capable of enhancing the workflow and problem-solving capabilities for reliability engineers by producing automatic equipment reporting and root-cause analysis through voice command.

GenAI can also help train new workers to operate complex equipment faster by making manuals and guides accessible through GenAI chat and search functionality.

  • Automated Computer Code Generation

Real-world examples: Autodesk/CloudNC CAM Assist, Microsoft/Rockwell Automation

GenAI can also be used in the generation of computer code to speed up programming tasks. Building GenAI capabilities into an industrial automation design environment can help engineers generate code using natural language prompts, automate routine tasks and improve design efficiency.

Manufacturing businesses can also work with technology providers to use GenAI to accelerate CAM programming time for industrial engineers. Automatically generating professional machining strategies at speed can drastically lower production hours for precision manufacturers.

  • HR Function Transformation

Real-world examples: Workday

Deploying GenAI streamlines HR operations within the manufacturing industry, enhancing capabilities in task automation, minimizing agent interactions, and refining onboarding and training processes, particularly important with high throughput of new staff and contractor workforce. This approach yields outcomes including substantial reductions in average handling time, noticeable improvements in user experience and the facilitation of additional outsourcing opportunities.

A&M recently supported a client, which had faced consecutive periods of shrinking margins, in an exercise to reduce costs by focusing on the optimization of the HR function leveraging GenAI.

Setting up for GenAI

Applying GenAI is entirely new for most enterprises, even to those already using AI and machine-learning (ML) innovations such as EdgeAI, Computer Vision and ML-based predictive maintenance. Because value-generating advancements in GenAI are progressing at pace in manufacturing, every enterprise should be preparing to apply this technology to develop competitive advantage.

From our experience supporting clients in starting their GenAI journey, here are some key considerations for setting up the foundations for GenAI:

  • Understanding the Opportunity - Establish a perspective of how GenAI fits into the broader AI strategy with a specific focus on the competitive opportunities and threats the functionality brings to the enterprise.
  • Establishing the Value Case: Evaluate at the outset the impact of investment into GenAI on various financial and non-financial metrics to gauge the extent of potential benefits. These include faster delivery time, enhanced employee productivity, customer experience and smoother operational processes. It is also important to keep scalability across divisions in mind while deploying such tools. Rapid prototyping is also a method which builds trust in the benefits case and tangibly demonstrates value to get wider buy in.
  • Identifying and Deploying the Platform: Choose the infrastructure and GenAI models and vendors that are the best fit for your business needs right from the start, whether it is cloud infrastructure provided by Microsoft, AWS and Google, or scalable models from innovative startups. When in need of GenAI solutions, consider whether buying is better than making. Software providers are moving to embed GenAI into key user journeys, rendering the need for in-house development unnecessary in some cases. At A&M, we are actively monitoring 80 software vendors (and growing) across 350 GenAI/AI functionality points and can see first-hand the tremendous pace of development from vendors.
  • Training and Knowledge Building: Create a sandbox environment to allow a selected group to experiment with the technology in a safe space. Establish a core set of GenAI domain experts who can help the organization understand, contextualize, scope and plan for applying GenAI at an economically appropriate pace. Training must encompass not just the how-tos, but also data security, data privacy and GenAI-specific considerations such as hallucinations (false or misleading information generated), model adoption and prompt injection.
  • Garbage In – Generative Mush Out: The outcome of any GenAI-assisted process will only be as good as the quality of the input data and the accuracy of prompts. It is therefore essential to maintain data integrity while deploying such tools.

GenAI is a rapidly progressing technology. Harnessed thoughtfully,it has the potential to unlock value within your business, enhancing human capabilities and driving growth.

Originally published 25 March 2024

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.